Data compression method, device, electronic device and storage medium

Abstract

The present application relates to a data compression method, device, electronic device, and storage medium, and relates to the field of wireless communication technology. The method includes: obtaining data to be compressed, and obtaining a compression bit width C, wherein the data to be compressed is data retransmitted each time under a hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the compression bit width C is determined based on the current hardware itself, and the compression bit width C is smaller than the effective bit width of the current hardware; converting the data to be compressed from a floating-point number to a fixed-point number to obtain a fixed-point number corresponding to the data to be compressed; and compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data. In this way, the fixed-point number corresponding to the data to be compressed can be compressed based on the compression bit width C, thereby reducing the storage space occupied by the data retransmitted each time under the hybrid automatic repeat request mechanism, thereby achieving the purpose of reducing hardware costs.

Description

Technical Field

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

[0002] New Radio (NR) and Long Term Evolution (LTE) wireless communication protocols often incorporate the Hybrid Automatic Repeat Request (HARQ) mechanism to ensure data accuracy. HARQ works by decoding a code block upon receipt and determining whether the decoding was successful. If the decoding fails, the transmitter retransmits the code block until the decoding is successful.

[0003] In related technologies, since each retransmitted data needs to be stored under the HARQ mechanism, it is easy to cause the data to occupy a large storage space. Therefore, how to compress the data to reduce the storage space occupied by the data has become a technical problem that needs to be solved urgently. Summary of the Invention

[0004] The present application provides a data compression method, device, electronic device and storage medium to solve the problem of large data storage space required by the HARQ mechanism in related technologies.

[0005] In a first aspect, an embodiment of the present application provides a data compression method, the method comprising:

[0006] Obtaining data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under a hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the compression bit width C is determined based on current hardware itself and is smaller than an effective bit width of the current hardware;

[0007] Converting the data to be compressed from floating point numbers to fixed point numbers to obtain fixed point numbers corresponding to the data to be compressed;

[0008] Based on the compression bit width C, the fixed-point number corresponding to the data to be compressed is compressed to obtain compressed data.

[0009] Optionally, compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data includes:

[0010] Divide the fixed-point number corresponding to the data to be compressed by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole;

[0011] The data occupying the bit width C in the high position of the calculated fixed-point number is deleted to obtain the compressed data.

[0012] Optionally, obtaining the data to be compressed includes:

[0013] Initial data carrying colored noise in a data channel is acquired, and the colored noise is converted into white noise to obtain the data to be compressed.

[0014] Optionally, converting the colored noise into white noise to obtain the data to be compressed includes:

[0015] Obtaining a covariance matrix corresponding to the colored noise, and decomposing the covariance matrix to obtain a lower triangular matrix, wherein the product of the lower triangular matrix and the conjugate transposed matrix of the lower triangular matrix is ​​equal to the covariance matrix;

[0016] Determine a target filter based on the lower triangular matrix, wherein filter coefficients of the target filter are an inverse matrix of the lower triangular matrix;

[0017] The colored noise is converted into white noise using the target filter to obtain the data to be compressed, wherein the values ​​of the data to be compressed are normally distributed.

[0018] Optionally, after compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data, the method further includes:

[0019] Storing the compressed data;

[0020] The stored compressed data is combined with the compressed data stored during the last retransmission under the hybrid automatic repeat mechanism, and the combined data is decoded and verified to determine whether the decoding is successful.

[0021] In a second aspect, an embodiment of the present application further provides a data compression device, the device comprising:

[0022] an acquisition module, configured to acquire data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under a hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the compression bit width C is determined based on current hardware itself and is smaller than an effective bit width of the current hardware;

[0023] A conversion module, configured to convert the data to be compressed from floating point numbers to fixed point numbers to obtain fixed point numbers corresponding to the data to be compressed;

[0024] The compression module is used to compress the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data.

[0025] Optionally, the compression module includes:

[0026] The calculation submodule is used to divide the fixed-point number corresponding to the compressed data by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole;

[0027] The deletion submodule is used to delete the data occupying the bit width C in the high position of the calculated fixed-point number to obtain the compressed data.

[0028] Optionally, the acquisition module includes:

[0029] The acquisition submodule is used to acquire initial data carrying colored noise in the data channel and convert the colored noise into white noise to obtain the data to be compressed.

[0030] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0031] Memory for storing computer programs;

[0032] The processor is configured to implement the data compression method described in the first aspect when executing a program stored in the memory.

[0033] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the data compression method as described in the first aspect.

[0034] The technical solution provided by the embodiment of the present application has the following advantages over the prior art: the method provided by the embodiment of the present application obtains data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under the hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the compression bit width C is determined by the current hardware itself and is smaller than the effective bit width of the current hardware; the data to be compressed is converted from a floating-point number to a fixed-point number to obtain a fixed-point number corresponding to the data to be compressed; and the fixed-point number corresponding to the data to be compressed is compressed based on the compression bit width C to obtain compressed data. In this way, the fixed-point number corresponding to the data to be compressed can be compressed based on the compression bit width C, thereby reducing the storage space occupied by the data retransmitted each time under the hybrid automatic repeat request mechanism, thereby reducing hardware costs; and because the data to be compressed has been subjected to a noise whitening operation, it conforms to the normal distribution characteristics, and thus can effectively reduce the problem of reduced decoding performance caused by reduced data accuracy due to compression, thereby ensuring that the decoding performance is not significantly affected. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0037] Figure 1 A flowchart of a data compression method provided in an embodiment of the present application;

[0038] Figure 2 A schematic diagram of a normal distribution provided in an embodiment of the present application;

[0039] Figure 3 A schematic structural diagram of a data compression device provided in an embodiment of the present application;

[0040] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0041] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0042] See also Figure 1 , Figure 1 A flow chart of a data compression method provided in an embodiment of the present application. Figure 1 As shown, the data compression method may include the following steps:

[0043] Step S102: Obtain the data to be compressed and the compression bit width C, wherein the data to be compressed is the data retransmitted each time under the hybrid automatic repeat mechanism, and the data to be compressed is a floating point number after performing a noise whitening operation. The compression bit width C is determined based on the current hardware itself and is smaller than the effective bit width of the current hardware.

[0044] Specifically, the compression bit width C refers to the bit width required to compress the data to be compressed under the hardware effective bit width. The size of the compression bit width C is determined by the current hardware's own conditions, and its value cannot be greater than or equal to the hardware effective bit width. For example, assuming that the hardware effective bit width is 8 bits, the compression bit width C can be any value between 1 and 7. The letter C here represents the compression bit width in this application. Of course, other letters can also be used to represent the compression bit width, which does not constitute a limitation of this application.

[0045] The data to be compressed is the data retransmitted under the hybrid automatic repeat request (HARQ) mechanism. The reason for compressing the data retransmitted under HARQ is that HARQ consumes a significant amount of memory. Taking LTE as an example, the maximum length of a transport block (TB) for single-layer transmission is 137,792 bits, while the length of a TB for quad-layer transmission can reach 502,624 bits. Each bit in each TB corresponds to a log-likelihood ratio (LLR). When hardware uses 8 bits to represent an LLR, storing a TB for single-layer transmission requires approximately 1MB of memory, while storing a TB for quad-layer transmission requires approximately 3.8MB of memory. Furthermore, compared to LTE, NR transmits larger amounts of data, and theoretically, a TB can have an unlimited size, thus requiring even greater memory space. It can be seen that when the hybrid automatic repeat request mechanism HARQ is performed under LTE or NR, a large amount of storage space is required, which not only affects the area required for hardware, but also reduces the memory access rate and throughput.

[0046] It should be noted that before compressing each retransmitted data under the hybrid automatic repeat request mechanism HARQ, it is necessary to perform a noise whitening operation on it, that is, convert the colored noise it carries into white noise so that it conforms to the normal distribution. In this way, the distribution of the compressed data in the range of [-1, +1] can reach about 68.27%, thereby effectively retaining the superposition influence of the Markov chain of the decimal part of most of the compressed data, effectively reducing the problem of data accuracy degradation caused by compression, and minimizing the decoding performance loss.

[0047] Step S104: convert the data to be compressed from floating point numbers to fixed point numbers to obtain fixed point numbers corresponding to the data to be compressed.

[0048] Since the data to be compressed is a floating-point number after noise whitening, it needs to be represented as a binary fixed-point number before compression. Take an LLR represented by an 8-bit fixed-point number as an example, where the first bit on the left is the highest bit, which represents the sign bit, and the other bits are data bits. When the floating-point number a is 0.11234, the floating-point number can be converted to a fixed-point number according to the following formula to obtain the fixed-point number a_fix corresponding to the floating-point number a: a_fix = b'10000000+0.11234*2 7 = b'10000000+b'1110= b'10001110, where b' represents the binary representation.

[0049] Step S106: Compress the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data.

[0050] Considering that the common combination of LLRs is an integer plus a fractional part, the integer part generally determines the positive or negative direction of the data, while the fractional part determines the impact of the Markov chain's superposition during decoding. For example, if LLR0 = +100000 is an error and the transport block (TB) length is 137792 bits, in traditional decoding, this means that unless all other codewords are correct, the Markov chain superposition of the fractional parts can correct this codeword. Therefore, retaining too much integer information does not significantly improve decoding performance. Therefore, when decoding data, the focus should be on the correct bits and fractional parts. Furthermore, after noise whitening, the data can be assumed to conform to a normal distribution. Based on the principle of normal distribution, the data distribution in the range [-1, +1] is approximately 68.27%, indicating that the data after noise whitening largely meets the decoding performance requirements. Therefore, when compressing the data to be compressed, the compression principle can be: retain the positive and negative signs of the data to be compressed, and perform corresponding compression on the decimal part. This can minimize the problem of decreased decoding performance caused by decreased data accuracy due to compression while ensuring the basic direction of decoding.

[0051] Through the above method, the fixed-point numbers corresponding to the data to be compressed can be compressed based on the compression bit width C, thereby reducing the storage space occupied by the data retransmitted each time under the hybrid automatic repeat mechanism, thereby achieving the purpose of reducing hardware costs; and because the data to be compressed has been subjected to a noise whitening operation, it conforms to the normal distribution characteristics, and thus can effectively reduce the problem of reduced decoding performance caused by the reduction in data accuracy due to compression, thereby ensuring that the decoding performance is not greatly affected.

[0052] In an optional embodiment, the above step S106, compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data, includes:

[0053] Divide the fixed-point number corresponding to the data to be compressed by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole;

[0054] The data occupying the bit width C in the high position of the calculated fixed-point number is deleted to obtain compressed data.

[0055] Specifically, when compressing the fixed-point number corresponding to the compressed data, the fixed-point number corresponding to the compressed data can be divided by 2. C, so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole. When the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole, the rightmost C-bit data of the fixed-point number corresponding to the data to be compressed is lost, and the leftmost C-bit data of the fixed-point number corresponding to the data to be compressed is replaced by 0. Next, the data occupying the bit width C in the high position of the calculated fixed-point number is deleted to obtain the compressed data. The data occupying the bit width C in the high position here refers to the data occupying the bit width C on the left, because in binary data representation, the highest bit is located at the leftmost side of the data and the lowest bit is located at the rightmost side of the data.

[0056] For example, when the floating-point number a is 0.11234, the floating-point number can be converted to a fixed-point number according to the following formula to obtain the fixed-point number a_fix corresponding to the floating-point number a: a_fix = b'10000000+0.11234*2 7 =b'10000000+b'1110=b'10001110. Assuming that the compression bit width C is 2, when compressing the floating-point number a, the binary fixed-point number b'10001110 can be first shifted right by 2 bits to obtain the binary fixed-point number b'00100011. Then, the highest 2 bits of the binary fixed-point number b'00100011 are deleted to obtain the binary fixed-point number b'100011, where the leftmost bit is the highest bit, indicating the sign of the data, and the other 5 bits represent the decimal part of the floating-point number. Converting the binary fixed-point number b'100011 to a floating-point number, we can get a'=+(b'11 / 2 5 ) = 0.09375. This shows that after compression, the floating-point number a suffers a data loss due to reduced precision of 0.11234 - 0.09375 = 0.01859. Although there is some data loss, the direction represented remains consistent and has little impact on overall decoding performance.

[0057] Through the above method, the sign bit of the data before and after compression can be kept unchanged, and only the decimal part is compressed accordingly. Therefore, under the premise of ensuring the basic direction of decoding, the problem of reduced decoding performance caused by reduced data accuracy due to compression can be minimized.

[0058] In an optional embodiment, the above step 102 of obtaining the data to be compressed includes:

[0059] Initial data with colored noise in the data channel is obtained, and the colored noise is converted into white noise to obtain data to be compressed.

[0060] Specifically, the colored noise mentioned above refers to the noise with uneven power distribution at different frequencies. Due to the interference of the colored noise, the data in the channel will not conform to the normal distribution. The white noise mentioned above refers to the noise with uniform power spectrum density in the entire frequency domain. Therefore, after the colored noise is converted into white noise, the data in the channel will conform to the normal distribution, such as Figure 2 shown.

[0061] Therefore, when obtaining the data to be compressed, the initial data carrying colored noise in the data channel can be obtained, and then the noise whitening operation is used to convert the colored noise into white noise to obtain the data to be compressed, so that the data to be compressed conforms to the normal distribution and can meet the decoding performance requirements after data compression.

[0062] In an optional embodiment, the above step of converting colored noise into white noise to obtain data to be compressed includes:

[0063] Obtain the covariance matrix corresponding to the colored noise and decompose the covariance matrix to obtain a lower triangular matrix, where the product of the lower triangular matrix and the conjugate transposed matrix of the lower triangular matrix is ​​equal to the covariance matrix;

[0064] Determine a target filter based on the lower triangular matrix, wherein the filter coefficients of the target filter are the inverse matrix of the lower triangular matrix;

[0065] The colored noise is converted into white noise using a target filter to obtain data to be compressed, wherein the values ​​of the data to be compressed are normally distributed.

[0066] Specifically, when converting colored noise into white noise, the covariance matrix corresponding to the colored noise can be obtained: , and decompose the covariance matrix to obtain the lower triangular matrix , where the product of the lower triangular matrix and the conjugate transposed matrix of the lower triangular matrix is ​​equal to the covariance matrix, satisfying the following formula:

[0067] ;

[0068] in, represents the covariance matrix corresponding to colored noise, represents the lower triangular matrix, and H represents the conjugate transposed matrix.

[0069] Then we can base it on the lower triangular matrix , determine the target filter, where the filter coefficient of the target filter is the inverse matrix of the lower triangular matrix, that is, the filter coefficient of the target filter is Then, the target filter can be used to convert the colored noise into white noise to obtain the data to be compressed. This can make the data to be compressed conform to the normal distribution and meet the decoding performance requirements after data compression.

[0070] In an optional embodiment, after compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C in step 106 to obtain compressed data, the method further includes:

[0071] Storing compressed data;

[0072] The stored compressed data is combined with the compressed data stored during the last retransmission under the hybrid automatic repeat mechanism, and the combined data is decoded and verified to determine whether the decoding is successful.

[0073] Specifically, after compressing the compressed data, the compressed data corresponding to the compressed data can also be stored. In this way, the stored compressed data can be combined with the compressed data stored during the last retransmission under the hybrid automatic retransmission mechanism, and the combined data can be decoded and verified to determine whether the decoding is successful, thereby realizing the automatic retransmission mechanism HARQ and ensuring the accuracy of the data.

[0074] In order to verify the compression performance loss of the data compression method provided in the embodiment of the present application, considering that 8 bits are commonly used in hardware to represent one LLR data, simulation tests are performed for different bit rate ranges, where the LLR is compressed to 6 bits, 4 bits, and 2 bits, respectively. The final decoding performance loss is shown in Table 1 below:

[0075]

[0076] Because different code rate ranges determine the code maps of the Low-Density Parity-Check Code (LDPC) used under the 3rd Generation Partnership Project (3GPP), different code maps have different impacts on decoding performance. Therefore, to demonstrate the compression effect of the present invention in different scenarios, it is necessary to cover as many different code maps as possible. The coding performance loss here refers to the loss value of the final decoded Block Error Ratio (BLER) mapped to the Signal-to-Noise Ratio (SNR).

[0077] As can be seen from the above table, when the data compression method provided by the embodiment of the present application is used for data compression, data accuracy will be lost, but since the data to be compressed conforms to the normal distribution, the data loss after compression can be controlled as low as possible within a low range to reduce the impact of the data loss after compression on the accuracy of the final decoding. As shown in the simulation results, when the data is compressed from 8 bits to 6 bits, the compression ratio is 25%, and the average loss of different code rate ranges is 0dB, but because the normal distribution is not perfect, the influence of noise causes jitter in some cases, so there will be a performance degradation of 0dB to 0.4dB under the worst loss. Similarly, when the data is compressed from 8 bits to 2 bits, the compression ratio is as high as 75%, and the normal distribution characteristics of the data are also affected more. Therefore, the average loss of different code rate ranges has a performance degradation of from 0.1dB to 1.3dB. As the compression coding code rate increases, the performance loss decreases, because the larger the data volume, the more its normal distribution characteristics can be guaranteed. Therefore, by utilizing the data compression method provided by the embodiment of the present application, data compression can be achieved through a smaller compression performance loss, reducing the demand for memory, thereby achieving the purpose of reducing hardware area and cost.

[0078] See also Figure 3 , Figure 3 This is a structural diagram of a data compression device provided in an embodiment of the present application. Figure 3 As shown, the data compression device 300 includes:

[0079] An acquisition module 302 is configured to acquire data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under the hybrid automatic repeat request mechanism and is a floating-point number after a noise whitening operation is performed. The compression bit width C is determined based on the current hardware itself and is smaller than the effective bit width of the current hardware.

[0080] The conversion module 304 is used to convert the floating point number to the fixed point number of the data to be compressed, so as to obtain the fixed point number corresponding to the data to be compressed;

[0081] The compression module 306 is configured to compress the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data.

[0082] Furthermore, the compression module 306 includes:

[0083] The calculation submodule is used to divide the fixed-point number corresponding to the compressed data by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole;

[0084] The deletion submodule is used to delete the data occupying the bit width C in the high position of the calculated fixed-point number to obtain compressed data.

[0085] Furthermore, the acquisition module 302 includes:

[0086] The acquisition submodule is used to acquire the initial data carrying colored noise in the data channel and convert the colored noise into white noise to obtain the data to be compressed.

[0087] Furthermore, the acquisition submodule includes:

[0088] an acquisition unit, configured to acquire a covariance matrix corresponding to the colored noise, and decompose the covariance matrix to obtain a lower triangular matrix, wherein the product of the lower triangular matrix and the conjugate transposed matrix of the lower triangular matrix is ​​equal to the covariance matrix;

[0089] a determining unit, configured to determine a target filter based on a lower triangular matrix, wherein a filter coefficient of the target filter is an inverse matrix of the lower triangular matrix;

[0090] The conversion unit is used to convert the colored noise into white noise by using a target filter to obtain data to be compressed, wherein the values ​​of the data to be compressed are normally distributed.

[0091] Furthermore, the data compression device 300 further includes:

[0092] A storage module, used for storing compressed data;

[0093] The combining module is used to combine the stored compressed data with the compressed data stored during the last retransmission under the hybrid automatic repeat mechanism, and decode and verify the combined data to determine whether the decoding is successful.

[0094] It should be noted that the data compression device 300 can implement the steps of the data compression method provided by any of the aforementioned method embodiments and can achieve the same technical effects, which will not be described in detail here.

[0095] like Figure 4 As shown, the embodiment of the present application further provides an electronic device, including a processor 411, a communication interface 412, a memory 413 and a communication bus 414, wherein the processor 411, the communication interface 412, and the memory 413 communicate with each other through the communication bus 414.

[0096] Memory 413, for storing computer programs;

[0097] In one embodiment of the present application, the processor 411 is configured to implement the steps of the data compression method provided in any one of the aforementioned method embodiments when executing the program stored in the memory 413 .

[0098] The electronic devices provided in the embodiments of the present application may specifically be modules capable of implementing communication functions or terminal devices containing such modules, and the terminal devices may be mobile terminals or smart terminals. A mobile terminal may specifically be at least one of a mobile phone, a tablet computer, and a laptop computer; a smart terminal may specifically be a terminal containing a wireless communication module, such as a smart car, a smart watch, a shared bicycle, or a smart cabinet; and a module may specifically be a wireless communication module, such as any one of a 2G communication module, a 3G communication module, a 4G communication module, a 5G communication module, and an NB-IOT communication module.

[0099] An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the data compression method provided in any of the aforementioned method embodiments are implemented.

[0100] The present application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the above-described data compression method.

[0101] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0102] The foregoing is merely a list of specific embodiments of the present application, intended to enable those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the broadest scope consistent with the principles and novel features of the present application.

Claims

1. A data compression method, characterized in that: The method comprises: Obtaining data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under a hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the data to be compressed includes at least a sign bit and a decimal part of an integer part, and the compression bit width C is determined based on current hardware itself and is smaller than an effective bit width of the current hardware; Converting the data to be compressed from floating point numbers to fixed point numbers to obtain fixed point numbers corresponding to the data to be compressed; Based on the compression bit width C, compressing the fixed-point number corresponding to the data to be compressed to obtain compressed data, wherein the compressed data retains the sign bit and the compressed decimal part of the data to be compressed; The step of compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data includes: Divide the fixed-point number corresponding to the data to be compressed by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole; The data occupying the bit width C in the high position of the calculated fixed-point number is deleted to obtain the compressed data.

2. The data compression method according to claim 1, wherein: The obtaining of the data to be compressed comprises: Initial data carrying colored noise in a data channel is acquired, and the colored noise is converted into white noise to obtain the data to be compressed.

3. The data compression method according to claim 2, wherein: The converting the colored noise into white noise to obtain the data to be compressed includes: Obtaining a covariance matrix corresponding to the colored noise, and decomposing the covariance matrix to obtain a lower triangular matrix, wherein the product of the lower triangular matrix and the conjugate transposed matrix of the lower triangular matrix is ​​equal to the covariance matrix; Determine a target filter based on the lower triangular matrix, wherein filter coefficients of the target filter are an inverse matrix of the lower triangular matrix; The colored noise is converted into white noise using the target filter to obtain the data to be compressed, wherein the values ​​of the data to be compressed are normally distributed.

4. The data compression method according to claim 1, wherein: After compressing the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data, the method further includes: Storing the compressed data; The stored compressed data is combined with the compressed data stored during the last retransmission under the hybrid automatic repeat mechanism, and the combined data is decoded and verified to determine whether the decoding is successful.

5. A data compression device, characterized in that: The device comprises: an acquisition module, configured to acquire data to be compressed and a compression bit width C, wherein the data to be compressed is data retransmitted each time under a hybrid automatic repeat request mechanism, and the data to be compressed is a floating-point number after a noise whitening operation is performed, and the data to be compressed includes at least a sign bit and a decimal part of an integer part, and the compression bit width C is determined based on current hardware itself and is smaller than the effective bit width of the current hardware; A conversion module, configured to convert the data to be compressed from floating point numbers to fixed point numbers to obtain fixed point numbers corresponding to the data to be compressed; a compression module, configured to compress the fixed-point number corresponding to the data to be compressed based on the compression bit width C to obtain compressed data, wherein the compressed data retains the sign bit and the compressed decimal part of the data to be compressed; Wherein, the compression module includes: The calculation submodule is used to divide the fixed-point number corresponding to the compressed data by 2 C , so that the fixed-point number corresponding to the data to be compressed is shifted right by C bits as a whole; The deletion submodule is used to delete the data occupying the bit width C in the high position of the calculated fixed-point number to obtain the compressed data.

6. The data compression device according to claim 5, characterized in that The acquisition module includes: The acquisition submodule is used to acquire initial data carrying colored noise in the data channel and convert the colored noise into white noise to obtain the data to be compressed.

7. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the data compression method according to any one of claims 1 to 4 when executing a program stored in a memory.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the data compression method according to any one of claims 1 to 4 is implemented.

Images