Data quantification method and device and storage medium

Abstract

The invention provides a data quantization method and device and a storage medium, and relates to the technical field of wireless communication, and the method comprises the steps: carrying out the digital down-conversion processing of an input signal based on a first quantization bit width, and obtaining a baseband signal; the first quantization bit width is determined based on the peak-to-average ratio of the input signal; performing channel estimation and equalization processing on the baseband signal based on a second quantization bit width; the second quantization bit width is obtained by real-time calculation based on the power of a pilot symbol in the baseband signal; demodulating the baseband signal after channel estimation and equalization processing, and performing dynamic bit width interception on soft information obtained by demodulation based on the second quantization bit width and the modulation mode to obtain soft information after interception; and carrying out saturation quantization interception processing on the intercepted soft information, and outputting the soft information to a decoder. Therefore, optimization of signal quantization precision and resource efficiency is realized.

Description

Technical Field

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

[0002] In single-carrier communication systems, especially in military wireless frequency-hopping systems, the frequency of each hop is different during the transmission process, so the frequency selective fading experienced by each hop is also different, resulting in different amplitudes of the received hop data at the receiving end; multi-carrier communication systems generally have a wider bandwidth and use orthogonal frequency division multiplexing (OFDM) + high-order dual modulation, which results in a high peak-to-average ratio (peak power to average power ratio) at the transmitting end of the multi-carrier communication system, that is, large amplitude fluctuations of the sample points in the time domain. Especially in OFDM multiple input multiple output (MIMO) systems, multipath fading is usually experienced, which will further lead to violent fluctuations in the amplitude of the time domain signal.

[0003] If the communication system uses amplitude modulation (with phase modulation) as the modulation mode, such as 8 Phase Shift Keying (8PSK), 16 Quadrature Amplitude Modulation (QAM) or higher-order 64QAM or 256QAM amplitude modulation, the amplitude is crucial to the entire baseband reception performance. If simple full-scale quantization is used, the entire reception processing data volume and cache will be relatively large. If the fixed amplitude quantization is used, the reception performance will be impaired. Summary of the invention

[0004] The present invention provides a data quantization method, device and storage medium, which are used to solve the defects of the prior art that the simple full-scale quantization has a large amount of data and cache for the entire receiving and processing, and the fixed amplitude quantization processing loses the receiving performance, so as to optimize the signal quantization accuracy and resource efficiency.

[0005] The present invention provides a data quantization method, which is applied to a communication system and comprises the following steps: Based on a first quantization bit width, performing digital down-conversion processing on the input signal to obtain a baseband signal; the first quantization bit width is determined based on a peak-to-average ratio of the input signal; Based on a second quantization bit width, channel estimation and equalization processing are performed on the baseband signal; the second quantization bit width is obtained by real-time calculation based on the power of the pilot symbol in the baseband signal; Demodulate the baseband signal after the channel estimation and equalization processing, and perform dynamic bit-width truncation on the demodulated soft information based on the second quantization bit-width and modulation method to obtain the truncated soft information; Perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0006] According to a data quantization method provided by the present invention, the determination method of the first quantization bit-width includes: Determine the first quantization bit-width according to the peak-to-average ratio of the input signal and the correspondence that 1 bit corresponds to 6 dB; The first quantization bit-width serves as the number of symbol bits of the input of the first-stage filter of digital downconversion; during the processing of the multi-stage filter of digital downconversion, the number of symbol bits of the input and output of each stage is the same.

[0007] According to a data quantization method provided by the present invention, the calculation method of the second quantization bit-width includes: Calculate the sum of the powers of the pilot set within a preset minimum scheduling unit; Calculate the second quantization bit-width based on the sum of the powers of the pilot set.

[0008] According to a data quantization method provided by the present invention, when the communication system is a single-carrier system, the minimum scheduling unit is a set of multiple pilot symbols and multiple data symbols; the sum of the powers of the pilot set is calculated in the time domain; When the communication system is a multi-carrier system, the minimum scheduling unit is a set of frequency-domain resource elements and user data resource elements; the sum of the powers of the pilot set is calculated in the frequency domain.

[0009] According to a data quantization method provided by the present invention, performing dynamic bit-width truncation on the demodulated soft information based on the second quantization bit-width and modulation method to obtain the truncated soft information includes: Based on the second quantization bit-width, determine the starting position for performing dynamic bit-width truncation on the demodulated soft information; Based on the modulation method, determine the third quantization bit-width for performing dynamic bit-width truncation on the demodulated soft information; Based on the starting position and the third quantization bit-width, perform dynamic bit-width truncation on the demodulated soft information to obtain the truncated soft information.

[0010] According to a data quantization method provided by the present invention, performing saturation quantization truncation processing on the truncated soft information includes: Merge the multiplexed truncated soft information and perform bit-width extension on the merged soft information; Based on the third quantization bit width, perform truncation processing on the soft information after bit width expansion.

[0011] The present invention also provides a data quantization device, which is applied to a communication system and includes the following modules: The first quantization module is used to perform digital down-conversion processing on the input signal based on the first quantization bit width to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average power ratio of the input signal; The second quantization module is used to perform channel estimation and equalization processing on the baseband signal based on the second quantization bit width; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal; The third quantization module is used to demodulate the baseband signal after the channel estimation and equalization processing, and perform dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method to obtain the truncated soft information; The fourth quantization module is used to perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0012] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the data quantization method as described in any one of the above.

[0013] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the data quantization method as described in any one of the above.

[0014] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the data quantization method as described in any one of the above.

[0015] The data quantization method, device, and storage medium provided by the present invention determine the first quantization bit width based on the peak-to-average power ratio of the input signal, perform digital down-conversion processing on the input signal according to the first quantization bit width, calculate the second quantization bit width in real time based on the power of the pilot symbols in the baseband signal, perform channel estimation and equalization processing on the baseband signal according to the second quantization bit width, perform dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method, and perform saturation quantization truncation processing on the truncated soft information, thereby optimizing the signal quantization accuracy and resource efficiency. Description of the Drawings

[0016] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0017] Figure 1 It is a schematic diagram of the processing flow of a communication receiving system provided by the related art.

[0018] Figure 2 It is a schematic diagram of the flow of the data quantization method provided by the present invention.

[0019] Figure 3 It is a schematic diagram of the overall flow of the communication system quantization embodiment provided by the present invention.

[0020] Figure 4 It is a schematic diagram of an example of maximum symbol bit quantization of DDC down-conversion based on the peak-to-average power ratio provided by the present invention.

[0021] Figure 5 It is a schematic diagram of the pilot bit width quantization estimation provided by the present invention.

[0022] Figure 6 It is a schematic diagram of the associated quantization of the average energy symbol bit and the modulation method provided by the present invention.

[0023] Figure 7 It is a schematic diagram of the saturation quantization of the decoding front end provided by the present invention.

[0024] Figure 8 It is a schematic diagram of the structure of the data quantization device provided by the present invention.

[0025] Figure 9 It is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed implementation manners

[0026] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0027] Figure 1 It is a schematic diagram of the processing flow of a communication receiving system provided by the related art, as Figure 1As shown in the figure, the typical communication receiving system includes radio frequency stage processing (including RF radio frequency processing and AD analog-to-digital sampling), symbol stage processing (including DDC digital down-conversion to baseband, FFT fast time-frequency conversion of multi-carrier systems, and time-frequency synchronization and subsequent channel estimation, equalization, and demodulation processing of multi-carrier and single-carrier systems), and bit-level processing (including descrambling, rate dematching, and decoding processing of the bit stream obtained after demodulation processing). The radio frequency analog part (i.e., the part between RF radio frequency processing and AD sampling) is usually analog automatic gain control (AGC) to ensure that the input signal amplitude is within a certain range.

[0028] Existing technologies usually focus on fixed-bitwidth quantization of data at a single node, and there are the following problems: ① The matching between the dynamic range and quantization accuracy is poor. When the peak-to-average ratio of the signal fluctuates, fixed quantization will cause signal truncation distortion or inefficient quantization; ② Resource waste. Fixed high-bit quantization leads to waste of power consumption and computing resources, and insufficient matching of modulation methods leads to resource waste; ③ Existing technologies only design at a single point on the receiving link and do not consider the digital domain link of the receiving system as a whole. For example, only consider the input of the decoding front end, that is, rate dematching combined quantization or hybrid automatic repeat request (HARQ) combined quantization; ④ The quantization parameters are based on statistics and belong to the category of non-real-time quantization technologies, and cannot truly reflect the true soft information amplitude of the current block to be decoded, thus reducing the decoding performance.

[0029] Similarly, some technologies use simple full-scale quantization for the entire link, then the entire receiving processing data volume and cache are relatively large. Some use fixed-amplitude quantization processing. When using different modulation methods, such as QPSK and 64QAM, the same quantization parameters are used, and the decoding performance of 64QAM will be greatly reduced.

[0030] Therefore, the present invention provides a data quantization method, device, and storage medium.

[0031] Figure 2 It is a schematic flowchart of the data quantization method provided by the present invention. As Figure 2 shown, this method is applied to a communication system and includes the following steps: Step 200: Perform digital down-conversion processing on the input signal based on the first quantization bitwidth to obtain a baseband signal; the first quantization bitwidth is determined based on the peak-to-average ratio of the input signal.

[0032] Step 201: Perform channel estimation and equalization processing on the baseband signal based on the second quantization bit width; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal.

[0033] Step 202: Demodulate the baseband signal after channel estimation and equalization processing, and perform dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method to obtain the truncated soft information.

[0034] Step 203: Perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0035] Specifically, the method provided by the present invention is applied to the receiving end of a communication system, and the signal processing accuracy and resource efficiency are optimized through a dynamic quantization strategy.

[0036] First, in the embodiments of the present invention, the first quantization bit width can be determined by the peak-to-average ratio (PAR) of the input signal. The input signal in the embodiments of the present invention refers to the digitalized intermediate-frequency or high-frequency signal obtained after AD analog-to-digital conversion at the receiving end of the communication system (such as 16-bit IQ sampling data after AD analog-to-digital conversion).

[0037] The input signal can be processed by digital down conversion (DDC) according to the first quantization bit width. DDC shifts the signal to the baseband through mixing and filtering, and reduces the sampling rate through a multi-stage decimation filter. The output of this step is the baseband signal, and its bit width is the same as the first quantization bit width.

[0038] Then, the second quantization bit width can be obtained by calculating the power of the pilot symbols in the baseband signal in real time. Then, based on the second quantization bit width, channel estimation and equalization processing are performed on the baseband signal. The channel estimation module calculates the channel response using the pilot symbols, and the equalization module corrects the data symbols according to the channel response.

[0039] It can be understood that since the channel conditions may change rapidly, the second quantization bit width is dynamically adjusted according to the power of the pilot symbols in the baseband signal, so as to ensure the accuracy of the intermediate calculation process of channel estimation and equalization processing, and at the same time avoid resource waste caused by high bit widths.

[0040] After obtaining the baseband signal after channel estimation and equalization processing, it can be demodulated, and the soft information is dynamically truncated according to the second quantization bit width and the modulation method. The demodulation process maps the baseband signal to soft information, and the bit width of the soft information can be adjusted in combination with the dynamically adjusted second quantization bit width and the complexity of the modulation method used for the received signal.

[0041] Finally, the truncated soft information can be subjected to saturation quantization truncation and output to the decoder. Saturation quantization prevents overflow by restricting the numerical range, while further compressing the data scale to meet the input requirements of the decoder. This operation should be based on the dynamic truncation of soft information to ensure that while reducing the data complexity, the key features of the soft information are maintained, thereby achieving efficient and reliable signal recovery in the decoding stage.

[0042] The data quantization method provided by the present invention determines the first quantization bit width based on the peak-to-average ratio of the input signal, performs digital down-conversion processing on the input signal according to the first quantization bit width, calculates the second quantization bit width in real time based on the power of the pilot symbols in the baseband signal, performs channel estimation and equalization processing on the baseband signal according to the second quantization bit width, and performs dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method, and performs saturation quantization truncation processing on the truncated soft information, thereby optimizing the signal quantization accuracy and resource efficiency.

[0043] According to a data quantization method provided by the present invention, the determination method of the first quantization bit width includes: Determine the first quantization bit width according to the peak-to-average ratio of the input signal and the correspondence of 1 bit corresponding to 6 dB; The first quantization bit width is used as the number of symbol bits of the input of the first-stage filter of digital down-conversion; during the processing of the multi-stage filter of digital down-conversion, the number of symbol bits of the input and output of each stage is the same.

[0044] Specifically, in the embodiments of the present invention, the first quantization bit width can be determined according to the correspondence of the peak-to-average ratio of the input signal and 1 bit corresponding to 6 dB. PAR represents the ratio of the peak power to the average power of the signal (unit: dB), and its value can reflect the dynamic range requirement of the signal.

[0045] For example, if the PAR of the input signal is 12 dB@0.01% (the probability that the ratio of the peak power to the average power dB = 12 dB appears is 0.01%), then according to the rule that 1 bit covers a 6 dB dynamic range, 2 bit symbol bits can be reserved (12 dB ÷ 6 dB / bit = 2 bit).

[0046] The process of determining the first quantization bit width is completed by statistically calculating the PAR value of the input signal in real time, ensuring that the quantization bit width can cover the peak amplitude of the signal while avoiding resource waste.

[0047] Subsequently, the first quantization bit width can be used as the number of input symbol bits of the first-stage filter of digital down-conversion. For example, if the input signal is 16-bit IQ data and its PAR is 12 dB, the input of the first-stage filter retains the highest 2 bit symbol bits (such as MSB[15:14] in 16 bits), and the remaining low bits are truncated.

[0048] In the embodiments of the present invention, during the multi-stage filter processing of digital down-conversion, the number of symbol bits of the input and output at each stage needs to be kept consistent. For example, the output of the first-stage filter may be extended to 33-40 bits due to multiplication and accumulation operations, but it needs to be truncated to 16 bits and retain the same 2-bit symbol bits as the input; subsequent-stage filters repeat this rule to ensure that the input and output of each stage use the same symbol bit width.

[0049] According to a data quantization method provided by the present invention, the calculation method of the second quantization bit width includes: Calculate the sum of the powers of the pilot set within a preset minimum scheduling unit; Calculate the second quantization bit width based on the sum of the powers of the pilot set.

[0050] Specifically, in the process of calculating the second quantization bit width, first, the sum of the powers of the pilot set can be calculated within a preset minimum scheduling unit. The minimum scheduling unit is the basic data unit processed by the receiving end and usually contains a group of pilot symbols and data symbols. The pilot set contains all pilot symbols. Pilot symbols are known reference signals, and their powers can directly reflect the current channel state.

[0051] In some embodiments, in the process of calculating the sum of powers, all pilot symbols within the minimum scheduling unit can be extracted; then, calculate the instantaneous power of the I / Q components of each pilot symbol (I 2 +Q 2 ); finally, accumulate the powers of all pilot symbols to obtain the total power sum (accu_power).

[0052] After calculating the sum of the powers of the pilot set, the second quantization bit width can be calculated according to this power sum. The second quantization bit width is used to guide the quantization truncation in the channel estimation and equalization process.

[0053] In some embodiments, the calculation method of calculating the second quantization bit width according to this power sum is: First, convert the total power sum (accu_power) to the average power (avg_power): avg_power = accu_power / N; where N represents the number of symbols included in the pilot set.

[0054] Then, calculate the amplitude (abs_mag): abs_mag = abs(avg_power); Finally, determine the quantization bit width (est_width) according to the amplitude (abs_mag): est_width = log2(abs_mag).

[0055] All intermediate quantization truncations in channel estimation and equalization calculations use est_width as a reference, dynamically scaling up and down to this bit width, thereby ensuring the quantization accuracy of the entire symbol processing while optimizing resource efficiency.

[0056] According to a data quantization method provided by the present invention, in the case where the communication system is a single-carrier system, the minimum scheduling unit is a set of multiple pilot symbols and multiple data symbols; the sum of the powers of the pilot set is calculated in the time domain; In the case where the communication system is a multi-carrier system, the minimum scheduling unit is a set of frequency-domain resource elements and user data resource elements; the sum of the powers of the pilot set is calculated in the frequency domain.

[0057] Specifically, different calculation methods for the sum of the powers of the pilot set can be designed according to the characteristics of the communication system to adapt to different channel estimation requirements.

[0058] In a single-carrier system, the minimum scheduling unit is composed of a set of multiple pilot symbols and multiple data symbols. Pilot symbols are usually designed in the time domain and are distributed at the front, middle, or tail of the symbols in the time domain symbols of the minimum scheduling unit, or in a combined distribution, and usually there is only one pilot port, so the power statistics of the pilot symbols are carried out in the time domain.

[0059] During the calculation process, the I / Q components of the time-domain pilot symbols are extracted, the instantaneous power of each symbol is calculated, and the powers of all pilot symbols are accumulated to obtain the total power sum. This time-domain statistical method is consistent with the continuity of the single-carrier signal and is convenient for real-time tracking of the channel state.

[0060] In a multi-carrier system, such as a Long Term Evolution (LTE) system, the minimum scheduling unit is composed of a set of frequency-domain resource elements (REs) and user data REs. Pilots are distributed in the frequency domain and the number of ports is configurable (1, 2, 4 ports). Based on the sparsity of the pilot symbols, port-based pilot power statistics are carried out to estimate the quantization bit width of each port, and the power statistics of the pilot symbols are carried out in the frequency domain.

[0061] During the calculation process, the frequency-domain I / Q values of each pilot RE are extracted, and their powers are calculated and accumulated. The frequency-domain statistical method matches the discrete sub-carrier structure of the multi-carrier signal and can accurately reflect the frequency-domain channel response.

[0062] According to a data quantization method provided by the present invention, based on the second quantization bit width and the modulation method, the demodulated soft information is dynamically truncated in bit width to obtain the truncated soft information, including: Determine the starting position for dynamically truncating the bit width of the demodulated soft information based on the second quantization bit width; Determine the third quantization bit width for dynamically truncating the bit width of the demodulated soft information based on the modulation method; Based on the starting position and the third quantization bit width, perform dynamic bit width truncation on the demodulated soft information to obtain the truncated soft information.

[0063] Specifically, in the process of dynamically truncating the bit width of the demodulated soft information according to the second quantization bit width and the modulation method, first, the starting position of the dynamic bit width truncation can be determined according to the second quantization bit width.

[0064] As mentioned above, the second quantization bit width is a quantization parameter calculated in real time through the pilot power (such as est_width = 12), and its value can indicate the high-order segment of the main energy distribution in the soft information. For example, if the second quantization bit width is 12, the truncation starting position of the 16-bit soft information is the 12th bit (such as MSB

[12] ), ensuring that the high-order segment covers the signal peak amplitude and avoiding truncation distortion.

[0065] Then, the third quantization bit width can be determined according to the modulation method.

[0066] The modulation method determines the density of the constellation points and the noise resistance ability, and can match different quantization precisions. Higher-order modulation (such as 256QAM) requires more quantization bits to distinguish dense constellation points; while lower-order modulation (such as QPSK) only requires a small number of bits to retain the key information of the symbol.

[0067] For example, if the modulation method is 256QAM and the second quantization bit width is 12, the third quantization bit width is 8 bits; if it is QPSK, the third quantization bit width is 2 bits.

[0068] This mapping relationship can be implemented through a preset rule table (for example, 256QAM → 8 bits, QPSK → 2 bits), ensuring that the quantization precision is consistent with the modulation complexity.

[0069] Finally, based on the starting position and the third quantization bit width, perform dynamic bit width truncation on the demodulated soft information to obtain the truncated soft information. The truncation range is intercepted downward from the starting position by the length of the third quantization bit width.

[0070] For example, if the starting position is 12 and the third quantization bit width is 8 bits, then intercept from the 12th bit to the 5th bit (a total of 8 bits, denoted as [12:5]); if the starting position is 12 and the third quantization bit width is 2 bits, then intercept from the 12th bit to the 11th bit ([12:11]).

[0071] A data quantization method provided by the present invention performs saturation quantization truncation processing on the truncated soft information, including: Merge the multiplexed truncated soft information, and perform bit-width expansion on the merged soft information; Based on the third quantization bit-width, perform truncation processing on the soft information with expanded bit-width.

[0072] Specifically, considering that there may be multiplexed soft information merging in the de-rate matching, saturation quantization truncation processing needs to be added in the embodiments of the present invention.

[0073] First, merge the multiplexed truncated soft information, and perform bit-width expansion on the merged soft information. The multiplexed soft information may come from HARQ retransmission merging or MIMO multi-stream detection (such as 4-way signal merging). Since the merging operation (such as summation) will cause the numerical range to expand, bit-width expansion is required to avoid overflow.

[0074] Subsequently, perform truncation processing on the expanded soft information based on the third quantization bit-width. The third quantization bit-width is usually determined by the previous modulation method (for example, 256QAM corresponds to 8 bits, and QPSK corresponds to 2 bits). Its function is to truncate the expanded high-bit-width data back to the target bit-width, and at the same time prevent overflow through saturation processing.

[0075] For example, for 16QAM 4-bit input soft information, the maximum number of 4-way soft information is soft0, soft1, soft2, and soft3 respectively. The sum (soft0 + soft1 + soft2 + soft3) is merged, expanded to 6 bits, and then 6-bit saturation truncated to 4 bits and output to the decoder input.

[0076] The data quantization method provided by the present invention is further elaborated below through embodiments in specific application scenarios.

[0077] Figure 3 It is a schematic diagram of the overall process of the communication system quantization embodiment provided by the present invention. As Figure 3 shown, this embodiment includes the following four parts: 1. Peak-to-average ratio-based DDC down-conversion maximum symbol bit quantization Figure 4 It is an example schematic diagram of the peak-to-average ratio-based DDC down-conversion maximum symbol bit quantization provided by the present invention. As Figure 4As shown, after AD analog-to-digital conversion, if it is zero intermediate frequency, the multi-stage DDC sampling rate is directly changed to the required baseband sampling rate. If it is real sampling, it first goes through quadrature mixing and then the multi-stage DDC sampling rate is changed to the required baseband sampling rate. During the process of changing the multi-stage DDC sampling rate, anti-aliasing filtering is required and shaping filtering is required for the last stage. The input and output of the filter need to be quantized, and the symbol bit number of the first-stage IQ is determined by the peak-to-average ratio of the input signal after sampling.

[0078] For example, if the estimated peak-to-average ratio of the input signal is 12 dB@0.01% (the ratio of peak power to average power dB = 12 dB appears with a probability of 0.01%), and the IQ bit width of the AD output is 16 bits, then the peak-to-average ratio of the signal is 12 dB, and the 1-bit quantization dB value is 6 dB. So the reserved quantization symbols for I or Q are 2 bits@16 bits, that is, the MSB[15:14] of 16-bit - 【15:0】 is the reserved symbol bit, which is the input of the first-stage mixing or anti-aliasing filter. Then the output I or Q bit width of the first-stage anti-aliasing filter (internal multiply-accumulate operation) is usually 33 bits to 40 bits (depending on the filter coefficients). The output of the first-stage filter is the input of the second-stage filter, and the input I or Q is often required to be no more than 16 bits. So the 33 bits to 40 bits need to be truncated and quantized to 16 bits. The quantization rule is to keep the reserved quantization symbols for I or Q of the input of the first-stage filter as 2 bits, that is, the reserved quantization symbols for I or Q of the input of the second-stage anti-aliasing filter are also 2 bits@16 bits. The input and output of the subsequent multi-stage filters follow the same rule.

[0079] Another example, if the peak-to-average ratio of the input signal after single-carrier sampling is 6 dB@0.01%, then the reserved quantization symbols for I or Q are 1 bit@16 bits, that is, the MSB

[15] of 16-bit - 【15:0】 is the reserved symbol bit, which is the input of the first-stage mixing or anti-aliasing filter. Similarly, the output of the first-stage filter is the input of the second-stage filter. So the 33 bits to 40 bits of the output need to be truncated and quantized to 16 bits. The quantization rule is to keep the reserved quantization symbols for I or Q of the input of the first-stage filter as 1 bit, that is, the reserved quantization symbols for I or Q of the input of the second-stage anti-aliasing filter are also 1 bit@16 bits. The input and output of the subsequent multi-stage filters follow the same rule.

[0080] 2. Channel Estimation and Equalization Calculation Quantization Based on Pilot Power Figure 5 It is a schematic diagram of pilot bit width quantization estimation provided by the present invention, as Figure 5As shown, for channel estimation and equalization calculation quantization based on pilot power, the pilot symbols in the received data are divided by port (or pilot samples), and the power of one or several groups of pilots belonging to the same port within the current scheduling granularity is statistically calculated. The quantization bit width is estimated in real time, and the intermediate calculation processes of channel estimation and equalization calculation are both subject to real-time dynamic quantization truncation processing with reference to this quantization bit width.

[0081] In a single-carrier system, pilots are usually designed in the time domain. In the time-domain symbols of the minimum scheduling unit, they are distributed at the front, middle, or tail of the symbol, or in a combined distribution, and usually there is only one pilot port. Therefore, the power statistics of pilot symbols are carried out in the time domain.

[0082] In a multi-carrier system, such as LTE, pilots are distributed in the frequency domain and the number of ports can be configured (1, 2, or 4 ports). Based on the sparsity of pilot symbols, pilot power statistics are carried out for each port, and the quantization bit width of each port is estimated. The power statistics of pilot symbols are carried out in the frequency domain.

[0083] First, the definition of the minimum scheduling unit is explained, that is, a data set of multiple time-domain symbols, which is the data input set for single decoding. For example, the single-carrier time-domain minimum scheduling unit includes pilots (multiple time-domain pilot symbols (samples)) and data (multiple data symbols (samples)). The pilot set is used for channel estimation, the data set is used as the equalization data input, the equalized output data set is then demodulated to obtain the soft information set, the soft information set is rate-matched and input to decoding, and finally the bit stream set is obtained through decoding. This bit stream set is the wireless communication information carried by the minimum scheduling unit.

[0084] Similarly, the multi-carrier minimum scheduling unit is similar. The difference is that the extraction of the pilot set in the multi-carrier system is in the frequency domain, and pilots are distributed on specific symbols in the frequency domain (the data or pilots within a frequency-domain symbol are also called REs (resource elements, which are different from the definition of time-domain symbols (samples))). After collecting the pilot REs based on ports in the multi-carrier system, it can be called the pilot set, and after collecting the user data REs based on users in the multi-carrier system, it can be called the user data set. Similarly, the pilot set is used for channel estimation, the data set is used as the equalization data input, the equalized output data set is then demodulated to obtain the soft information set, the soft information set is rate-matched and input to decoding, and finally the bit stream set is obtained through decoding. This bit stream set is the wireless communication information carried by the minimum scheduling unit.

[0085] Pilot power estimation and quantization bit width calculation: For multiple IQs (I_0 + Q_0 1j, I_1 + Q_1 1j, I_2 + Q_2 1j, …, I_n - 1 + Q_n - 1 1j) Calculate the power sum of sample points and / or the power sum of resource REs, that is, accu_power = sum(I_0^2 + Q_0^2 + I_1^2 + Q_1^2 + I_2^2 + Q_2^2 + … + I_n-1^2 + Q_n-1^2), average power = avg_power = accu_power / N, amplitude abs_mag = abs(avg_power), calculate the quantization bit width = est_width = log2(abs_mag), and the quantization bit width range is 1 to 16.

[0086] In the channel estimation and equalization calculations, all intermediate quantization truncations use est_width as a reference, dynamically scaling to this bit width up and down, thereby ensuring the quantization accuracy of the entire symbol processing while optimizing resource efficiency.

[0087] 3. Quantization associated with the average energy symbol bit and modulation mode Figure 6 is a schematic diagram of the quantization associated with the average energy symbol bit and modulation mode provided by the present invention. As Figure 6 shown, there is already est_width parameter information at the previous stage. The average energy symbol bit can be determined according to the est_width parameter. Here, first demodulate the data output by the equalization and convert it into 16-bit soft information. For example, when the current est_width = 12 and the modulation mode is: 256QAM, dynamically quantize it to 8 bits, and intercept the 8-bit output truncated soft information from [15:0] to [12:5]; 64QAM, dynamically quantize it to 6 bits, and intercept the 6-bit output truncated soft information from [15:0] to [12:7]; 16QAM, dynamically quantize it to 4 bits, and intercept the 4-bit output truncated soft information from [15:0] to [12:9]; 8PSK, dynamically quantize it to 3 bits, and intercept the 3-bit output truncated soft information from [15:0] to [12:10]; QPSK and BPSK, dynamically quantize it to 2 bits, and intercept the 2-bit output truncated soft information from [15:0] to [12:11].

[0088] 4. Saturation quantization at the decoding front end Figure 7 is a schematic diagram of the saturation quantization at the decoding front end provided by the present invention. As Figure 7As shown, considering that there may be multiple soft information merges in the solution rate matching, that is, soft information with different precisions (8bit / 6bit / 4bit / 2bit) enters the input buffer (BUF) and is merged through the circular BUF. Here, saturation quantization truncation processing must be added. For example, for 16QAM 4bit input soft information, a maximum of 4-way merge, sum(soft0 + soft1 + soft2 + soft3), is extended to 6bit, and then 6bit is saturated and truncated to 4bit for output to the decoder input.

[0089] The data quantization device provided by the present invention will be described below. The data quantization device described below can be correspondingly referred to the data quantization method described above.

[0090] Figure 8 is a schematic structural diagram of the data quantization device provided by the present invention. As Figure 8 shown, this device is applied to a communication system and includes the following modules: The first quantization module 800 is used to perform digital down-conversion processing on the input signal based on the first quantization bit width to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average ratio of the input signal. The second quantization module 810 is used to perform channel estimation and equalization processing on the baseband signal based on the second quantization bit width; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal. The third quantization module 820 is used to demodulate the baseband signal after channel estimation and equalization processing, and perform dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method to obtain truncated soft information. The fourth quantization module 830 is used to perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0091] According to a data quantization device provided by the present invention, the determination method of the first quantization bit width includes: Determine the first quantization bit width according to the peak-to-average ratio of the input signal and the correspondence of 1 bit corresponding to 6dB. The first quantization bit width is used as the number of symbol bits of the input of the first-stage filter of digital down-conversion; during the processing of the multi-stage filter of digital down-conversion, the number of symbol bits of the input and output of each stage is the same.

[0092] According to a data quantization device provided by the present invention, the calculation method of the second quantization bit width includes: Calculate the sum of the powers of the pilot set within a preset minimum scheduling unit. Calculate the second quantization bit width based on the sum of the powers of the pilot set.

[0093] According to a data quantization device provided by the present invention, in the case where the communication system is a single-carrier system, the minimum scheduling unit is a set of multiple pilot symbols and multiple data symbols; the sum of the powers of the pilot set is calculated in the time domain; In the case where the communication system is a multi-carrier system, the minimum scheduling unit is a set of frequency-domain resource elements and user data resource elements; the sum of the powers of the pilot set is calculated in the frequency domain.

[0094] According to a data quantization device provided by the present invention, based on a second quantization bit width and a modulation mode, dynamic bit-width truncation is performed on the demodulated soft information to obtain truncated soft information, including: Based on the second quantization bit width, determine the starting position for performing dynamic bit-width truncation on the demodulated soft information; Based on the modulation mode, determine a third quantization bit width for performing dynamic bit-width truncation on the demodulated soft information; Based on the starting position and the third quantization bit width, perform dynamic bit-width truncation on the demodulated soft information to obtain truncated soft information.

[0095] According to a data quantization device provided by the present invention, perform saturation quantization truncation processing on the truncated soft information, including: Merge the multiplexed truncated soft information and perform bit-width expansion on the merged soft information; Based on the third quantization bit width, perform truncation processing on the soft information after bit-width expansion.

[0096] Figure 9 It is a schematic structural diagram of an electronic device provided by the present invention. As Figure 9 shown, the electronic device may include: a processor 910, a communication interface 920, a memory 930, and a communication bus 940. Among them, the processor 910, the communication interface 920, and the memory 930 complete communication with each other through the communication bus 940. The processor 910 may call logical instructions in the memory 930 to execute a data quantization method, and the method includes: Based on a first quantization bit width, perform digital down-conversion processing on the input signal to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average ratio of the input signal; Based on a second quantization bit width, perform channel estimation and equalization processing on the baseband signal; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal; Demodulate the baseband signal after channel estimation and equalization processing, and based on the second quantization bit width and the modulation mode, perform dynamic bit-width truncation on the demodulated soft information to obtain truncated soft information; Perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0097] In addition, when the logic instructions in the above-mentioned memory 930 can be implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0098] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the data quantization method provided by the above-mentioned various methods. The method includes: Based on a first quantization bit width, perform digital down-conversion processing on the input signal to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average power ratio of the input signal; Based on a second quantization bit width, perform channel estimation and equalization processing on the baseband signal; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal; Demodulate the baseband signal after channel estimation and equalization processing, and perform dynamic bit width truncation on the demodulated soft information based on the second quantization bit width and the modulation method to obtain truncated soft information; Perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0099] On yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the data quantization method provided by the above-mentioned various methods. The method includes: Based on a first quantization bit width, perform digital down-conversion processing on the input signal to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average power ratio of the input signal; Based on a second quantization bit width, perform channel estimation and equalization processing on the baseband signal; the second quantization bit width is calculated in real time based on the power of the pilot symbols in the baseband signal; Demodulate the baseband signal after channel estimation and equalization processing, and perform dynamic bit-width truncation on the soft information obtained by demodulation based on the second quantization bit-width and modulation method to obtain the truncated soft information; Perform saturation quantization truncation processing on the truncated soft information and output it to the decoder.

[0100] The device embodiments described above are merely illustrative. 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, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0101] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment 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 solutions, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A data quantization method, characterized in that, Applied to a communication system, including: Performing digital down-conversion processing on an input signal based on a first quantization bit width to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average power ratio of the input signal; Performing channel estimation and equalization processing on the baseband signal based on a second quantization bit width; the second quantization bit width is calculated in real time based on the power of pilot symbols in the baseband signal; Demodulating the baseband signal after the channel estimation and equalization processing, and dynamically truncating the bit width of the demodulated soft information based on the second quantization bit width and the modulation method to obtain truncated soft information; Performing saturation quantization truncation processing on the truncated soft information and outputting it to a decoder.

2. The data quantization method according to claim 1, wherein The determination method of the first quantization bit width includes: Determining the first quantization bit width according to the peak-to-average power ratio of the input signal and the correspondence of 1 bit corresponding to 6 dB; The first quantization bit width serves as the number of symbol bits of the input of the first-stage filter of digital down-conversion; during the processing of the multi-stage filter of digital down-conversion, the number of symbol bits of the input and output of each stage is the same.

3. The data quantization method according to claim 1, wherein The calculation method of the second quantization bit width includes: Calculating the sum of the powers of the pilot set within a preset minimum scheduling unit; Calculating the second quantization bit width based on the sum of the powers of the pilot set.

4. The data quantization method according to claim 3, wherein In the case where the communication system is a single-carrier system, the minimum scheduling unit is a set of multiple pilot symbols and multiple data symbols; the sum of the powers of the pilot set is calculated in the time domain; In the case where the communication system is a multi-carrier system, the minimum scheduling unit is a set of frequency-domain resource elements and user data resource elements; the sum of the powers of the pilot set is calculated in the frequency domain.

5. The data quantization method according to claim 1, 3 or 4, characterized in that Dynamically truncating the bit width of the demodulated soft information based on the second quantization bit width and the modulation method to obtain truncated soft information, including: Determining the starting position for dynamically truncating the bit width of the demodulated soft information based on the second quantization bit width; Determining a third quantization bit width for dynamically truncating the bit width of the demodulated soft information based on the modulation method; Dynamically truncating the bit width of the demodulated soft information based on the starting position and the third quantization bit width to obtain truncated soft information.

6. The data quantization method according to claim 5, wherein Performing saturation quantization truncation processing on the truncated soft information, including: Merging the multiplexed truncated soft information and expanding the bit width of the merged soft information; Performing truncation processing on the soft information with the expanded bit width based on the third quantization bit width.

7. A data quantization device, characterized in that, Applied to a communication system, including: A first quantization module for performing digital down-conversion processing on an input signal based on a first quantization bit width to obtain a baseband signal; the first quantization bit width is determined based on the peak-to-average power ratio of the input signal; A second quantization module for performing channel estimation and equalization processing on the baseband signal based on a second quantization bit width; the second quantization bit width is calculated in real time based on the power of pilot symbols in the baseband signal; A third quantization module for demodulating the baseband signal after the channel estimation and equalization processing, and dynamically truncating the bit width of the demodulated soft information based on the second quantization bit width and the modulation method to obtain truncated soft information; A fourth quantization module, configured to perform saturation quantization truncation processing on the truncated soft information and output the result to a decoder.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the data quantization method according to any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by a processor, it implements the data quantization method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the data quantization method according to any one of claims 1 to 6.

Images