Simulation filter unevenness compensation method and system

Abstract

The invention provides an analog filter unevenness compensation method and system, and the method comprises the steps: carrying out the full-load scheduling of a known modulation symbol on a channel bandwidth, forming first frequency domain data used for generating calibration data, converting the first frequency domain data to a time domain for timing deviation compensation, and obtaining the calibration data on the frequency domain of an analog filter, and compensating the service data in the frequency domain according to the calibration data. Known modulation symbols are dispatched on channel bandwidth in a full-load mode, calibration data on a frequency domain are obtained after timing deviation estimation, and then the calibration data can be used for directly conducting unevenness compensation on the frequency domain according to subcarriers. The method can well reflect the passband amplitude and phase unevenness characteristics of the analog filter in the frequency domain, and reduces the operation complexity at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of filter calibration, and in particular to a method and system for compensating for unevenness of an analog filter. Background Art

[0002] Analog filter design is relatively mature, with several typical models including Butterworth and Chebyshev filters. These filters each have their own unique characteristics, allowing for compromises to be made based on specific requirements. However, regardless of the type of filter, they all exhibit passband amplitude and phase unevenness. In particular, the passband edge exhibits significant amplitude and phase differences compared to the midband. Furthermore, amplitude and phase jitter also exists within the passband. In OFDM systems, the effects of analog filter passband amplitude and phase unevenness are transferred to the digital baseband. In the frequency domain, different subcarriers correspond to different amplitudes and phases, which degrades transceiver performance and specifications. Therefore, calibration and compensation are necessary.

[0003] Existing technology addresses the issue of amplitude and phase unevenness in the analog filter passband by adding a first-order FIR filter and a first-order IIR filter to the digital baseband. The former addresses amplitude unevenness, while the latter addresses phase unevenness. The orders of both filters are pre-constrained, and the filter coefficients are configurable. This approach uses the frequency-domain characteristics of the two filters to fit the inverse characteristics of the analog filter's passband amplitude and phase to achieve unevenness compensation. To reduce computational complexity, existing technologies limit the order of FIR and IIR filters to a small value. However, this approach fails to accurately fit the inverse characteristics of the analog filter's passband amplitude and phase, resulting in poor compensation performance. Furthermore, generating FIR and IIR filter coefficients is challenging: first, the inverse characteristics of the amplitude and phase within the analog filter's passband must be obtained. However, these cannot be obtained for the transition band and stopband, requiring the use of passband edge results or zero values instead. Furthermore, the amplitude must be assumed to be symmetric about the DC subcarrier. Secondly, after converting the frequency-domain inverse characteristics to the time domain, the time-domain results must be truncated to filter coefficients of the corresponding order length, which affects the original frequency-domain inverse characteristics. Summary of the Invention

[0004] Based on the above description, the present invention provides a method and system for compensating for unevenness of an analog filter, aiming to solve the technical problem of poor compensation performance and complex calculation of analog filters in the prior art.

[0005] A method for compensating for unevenness of an analog filter, comprising:

[0006] Step A1: fully schedule known modulation symbols on the channel bandwidth to form first frequency domain data;

[0007] Step A2, converting the first frequency domain data into the time domain to obtain first time domain data;

[0008] Step A3: sending first time domain data, and obtaining time domain received data through radio frequency processing;

[0009] Step A4, performing timing offset compensation on the time domain received data according to the first time domain data to obtain second time domain data;

[0010] Step A5, converting the second time domain data into the frequency domain to form second frequency domain data;

[0011] Step A6, obtaining calibration data of the analog filter based on the first frequency domain data and the second frequency domain data;

[0012] Step A7: receiving service data in the time domain and converting it into the frequency domain, compensating the service data in the frequency domain according to the calibration data, and then outputting the result.

[0013] Furthermore, step A4 includes:

[0014] Step A41: calculating a maximum correlation value based on the first time domain data and the time domain received data as a first maximum correlation value, obtaining a first index number associated with the first maximum correlation value, and obtaining an index number preceding the first index number as a second index number;

[0015] Step A42: extracting data of a preset length from the time-domain received data starting from a preset number of sampling points in advance based on the first index number to form first target data, and extracting data of a preset length starting from a preset number of sampling points in advance based on the second index number to form second target data;

[0016] Step A43, calculating the correlation value of the first target data as the second correlation value, calculating the maximum correlation value of the first output result after the first target data is filtered by the first filtering module as the third maximum correlation value, and calculating the maximum correlation value of the second output result after the second target data is filtered by the second filtering module as the fourth maximum correlation value;

[0017] Step A44: Determine the numerical relationship among the second correlation value, the third maximum correlation value, and the fourth maximum correlation value:

[0018] If the second correlation value is the largest, execute step A45;

[0019] If the third maximum correlation value is the largest, execute step A46;

[0020] If the fourth maximum correlation value is the largest, execute step A47;

[0021] Step A45, using the first index number to perform integer timing offset compensation on the time domain received data;

[0022] Step A46, performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a set of filters associated with the first index number and the third maximum correlation value in the first filtering module;

[0023] Step A47, performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a set of filters associated with the second index number and the fourth maximum correlation value in the second filtering module;

[0024] The first filter module and the second filter module both include a plurality of filter groups, and the preset number of sampling points and the preset length are associated with the number of coefficients of each filter group.

[0025] Furthermore, step A41 includes:

[0026] Step A411: record a complete known modulation symbol in the first time domain data as x(n), n=0, 1, ..., corr len -1 is padded with 0 at the end to a length of N and then transformed into the frequency domain. The conjugate of the complex number is then taken to obtain the first frequency domain result. At the same time, n is changed from 0 to corr len +2×offset len -1 part of the time domain received data y(n), n=0,1,...,corr len +2×offset len -1 is padded with 0 at the end to a length of N and then transformed into the frequency domain to obtain the second frequency domain result;

[0027] Step A412: multiply the first frequency domain result and the second frequency domain result to obtain a third frequency domain result;

[0028] Step A413: convert the third frequency domain result to the time domain to obtain the first time domain result, and then retain the first 2×offset in the first time domain result. len +1 data as the first relevant value;

[0029] Step A414: taking the first correlation value corresponding to the maximum absolute value as the first maximum correlation value, and taking the index number of the first maximum correlation value in the first correlation value as the first index number;

[0030] Among them, corr len is the associated length; offset len The preset offset points.

[0031] Furthermore, in step A42, the number of sampling points is rounded down. The default length is corr len +N tap ;

[0032] Among them, N tap is the number of coefficients in each filter group.

[0033] Furthermore, in step A42, the first target data extracted is y0(n), n=0, 1, ..., corr len +N tap -1;

[0034] In step A43, the second correlation value is calculated using the following formula:

[0035]

[0036] Here, corr_try0 is the second correlation value.

[0037] Furthermore, in step A43, the first filter module includes N flt The first target data passes through each filter group in the first filter module and discards N tap filter outputs to obtain a first output result;

[0038] In step A43, the third correlation value of the first output result is calculated using the following formula:

[0039]

[0040] Among them, corr_try_tmp1 is the third correlation value;

[0041] yf0(n),n=0,1,...,corr len -1 is the first output result;

[0042] In step A43, the maximum value among the third correlation values is found as the third maximum correlation value.

[0043] Furthermore, in step A43, the second filter module includes N flt The second target data passes through each filter group in the second filter module and discards N tap filter output to obtain a second output result;

[0044] In step A43, the fourth correlation value of the second output result is calculated using the following formula:

[0045]

[0046] Wherein, corr_try_tmp2 is the fourth correlation value;

[0047] yf1(n),n=0,1,...,corr len -1 is the second output result;

[0048] In step A43, the maximum value among the fourth correlation values is found as the fourth maximum correlation value.

[0049] Furthermore, in step A6, the calculation formula of the calibration data is as follows:

[0050]

[0051] in, is the calibration data;

[0052] L(k), k=0,1,...,K-1 is the first frequency domain data;

[0053] Y(k), k=0,1,...,K-1 is the second frequency domain data;

[0054] N step is the preset number of subcarriers;

[0055] K is the total number of subcarriers within the channel bandwidth.

[0056] Furthermore, in step A7, the calculation formula for compensating the service data in the frequency domain according to the calibration data is as follows:

[0057]

[0058] Among them, Y r (k), k=0,1,...,K-1 is the service data in the frequency domain;

[0059] Y comp (k) is the service data after calibration and compensation of the service data in the frequency domain.

[0060] An analog filter unevenness compensation system is used to execute the above analog filter unevenness compensation method.

[0061] The beneficial technical effect of the present invention is that by fully scheduling known modulation symbols on the channel bandwidth, calibration data in the frequency domain is obtained after timing deviation estimation, and then the calibration data can be used to directly perform unevenness compensation according to subcarriers in the frequency domain, which can better reflect the analog filter passband amplitude and phase unevenness characteristics in the frequency domain while reducing the computational complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 and Figure 3 This is a flowchart of a method for compensating for unevenness of an analog filter according to the present invention;

[0063] Figure 2 A flowchart illustrating a method for compensating for unevenness of an analog filter according to the present invention;

[0064] Figure 4 This is a flowchart of integer timing deviation acquisition of an analog filter unevenness compensation method of the present invention;

[0065] Figure 5 This is a flow chart of obtaining fractional timing deviation in an analog filter unevenness compensation method according to the present invention. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0067] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other.

[0068] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0069] See also Figure 1-2 The present invention provides a method for compensating for unevenness of an analog filter, comprising:

[0070] Step A1: fully schedule known modulation symbols on the channel bandwidth to form first frequency domain data;

[0071] Step A2, converting the first frequency domain data into the time domain to obtain first time domain data;

[0072] Step A3: sending first time domain data, and obtaining time domain received data through radio frequency processing;

[0073] Step A4, performing timing offset compensation on the time domain received data according to the first time domain data to obtain second time domain data;

[0074] Step A5, converting the second time domain data into the frequency domain to form second frequency domain data;

[0075] Step A6, obtaining calibration data of the analog filter based on the first frequency domain data and the second frequency domain data;

[0076] Step A7: receiving service data in the time domain and converting it into the frequency domain, compensating the service data in the frequency domain according to the calibration data, and then outputting the result.

[0077] In step A1 , all subcarriers within the channel bandwidth carry known modulation symbols L(k), k=0, 1, ..., K-1, where K is the total number of subcarriers within the channel bandwidth.

[0078] In step A2, according to the OFDM system sending process, the frequency domain modulation symbol, ie, the first frequency domain data, is subjected to inverse fast Fourier transform IFFT, and a cyclic prefix CP is added to obtain known time domain data, ie, the first time domain data, and sent out.

[0079] In step A3, after receiving the first time domain data after RF processing at the transmitting end and the receiving end, the first time domain data is affected by the amplitude and phase unevenness of the analog filter passband. Next, digital baseband processing is performed on the transmitted first time domain data and the received time domain data.

[0080] See also Figure 3 , further, step A4 includes:

[0081] Step A41: calculating a maximum correlation value based on the first time domain data and the time domain received data as a first maximum correlation value, obtaining an index number associated with the first maximum correlation value as a first index number, and obtaining an index number previous to the first index number as a second index number;

[0082] Step A42: extracting data of a preset length from the time-domain received data starting from a preset number of sampling points in advance based on the first index number to form first target data, and extracting data of a preset length starting from a preset number of sampling points in advance based on the second index number to form second target data;

[0083] Step A43, calculating the correlation value of the first target data as the second correlation value, calculating the maximum correlation value of the first output result after the first target data is filtered by the first filtering module as the third maximum correlation value, and calculating the maximum correlation value of the second output result after the second target data is filtered by the second filtering module as the fourth maximum correlation value;

[0084] Step A44: Determine the numerical relationship between the second correlation value, the third maximum correlation value, and the fourth maximum correlation value:

[0085] If the second correlation value is the largest, execute step A45;

[0086] If the third maximum correlation value is the largest, execute step A46;

[0087] If the fourth maximum correlation value is the largest, execute step A47;

[0088] Step A45, using the first index number to perform integer timing offset compensation on the time domain received data;

[0089] Step A46, performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a set of filters associated with the first index number and the third maximum correlation value in the first filtering module;

[0090] Step A47, performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a set of filters associated with the second index number and the fourth maximum correlation value in the second filtering module;

[0091] The first filter module and the second filter module both include a plurality of filter groups, and the preset number of sampling points and the preset length are associated with the number of coefficients of each filter group. The filter is an interpolation filter.

[0092] See also Figure 4-5 , further, step A41 includes:

[0093] Step A411: Let a complete OFDM symbol of the first time domain data (excluding the cyclic prefix CP) be recorded as: x(n), n = 0, 1, ..., corr len -1, corr len is the association length; it is offset in advance of the start of a complete OFDM symbol (excluding the cyclic prefix CP) of the currently considered received time domain data. len +res len Sampling points for data reception, where offset len Indicates the preset offset points, res len Represents the number of redundant sampling points, and the received time domain data is recorded as:

[0094] y(n),n=-res len ,-res len +1,...,corr len +2×offset len -1 ；

[0095] Set x(n),n=0,1,...,corr len -1 is padded with 0 at the end to a length of N, and then an N-point fast Fourier transform (FFT) is performed to the frequency domain. Then the conjugate of the complex number is calculated to obtain the first frequency domain result. At the same time, the part y(n), n=0,1,...,corr len +2×offset len The -1 part is padded with 0 at the end to a length of N, and then an N-point fast Fourier transform (FFT) is performed to the frequency domain to obtain the second frequency domain result;

[0096] Step A412: multiply the first frequency domain result and the second frequency domain result to obtain a third frequency domain result;

[0097] Step A413: Perform an N-point inverse fast Fourier transform (IFFT) on the third frequency domain result to convert it to the time domain to obtain a first time domain result, and retain the first 2×offset in the first time domain result.len +1 data to get the first correlation value of the series;

[0098] Step A414: After taking the absolute value abs of each first correlation value, the first correlation value corresponding to the maximum absolute value is taken as the first maximum correlation value, and the index number (correlation value number) corresponding to the first maximum correlation value is recorded as the first index number as idx. The value range of idx is 0 to 2×offset. len .

[0099] Furthermore, the first filtering module and the second filtering module both include N flt Groups of filters, each group of filters includes N tap coefficients, through which the fractional timing deviation can be interpolated as data.

[0100] Furthermore, in step A42, the number of sampling points is rounded down. The default length is corr len +N tap ;

[0101] Among them, N tap is the number of coefficients in each filter group.

[0102] Round down in advance according to the first index number idx sampling points, extract the length corr from the time domain received data y(n). len +N tap The data is recorded as: y0(n), n=0,1,...,corr len +N tap -1, namely the first target data;

[0103] In step A43, the second correlation value is calculated according to the following calculation formula:

[0104]

[0105] Here, corr_try0 is the second correlation value.

[0106] Furthermore, in step A43, the first target data y0(n), n=0, 1, ..., corr len +N tap -1 respectively pass through N in the first filtering module flt Group of filters, the filter is FIR filter, each group of filters loses N tap filter output, retain the next corr len filter output, that is, the first output result is expressed as: yf0(n), n=0,1,...,corr len-1; the calculation formula of the third correlation value is as follows:

[0107]

[0108] Among them, corr_try1_tmp1 is the third correlation value;

[0109] yf0(n),n=0,1,...,corr len -1 is the first output result.

[0110] N flt The filter gets N flt The third correlation value is found from corr_try_tmp1, that is, corr_try1 is used as the third maximum correlation value. The corresponding third index number (the number of each filter group of the first filtering module) is frac_try1, and the value ranges from 1 to N flt .

[0111] In step A42, the first index number is rounded down in advance by one index number idx-1. sampling points, extract the length corr from the time domain received data y(n). len +N tap The data is recorded as: y1(n), n=0,1,...,corr len +N tap -1, that is, the second target data; then y1(n), n=0,1,...,corr len +N tap -1 respectively pass through N in the second filtering module flt Group of filters, the filter is FIR filter, each group of filters loses N tap filter output, retain the next corr len filter output, that is, the second output result is: yf1(n), n=0,1,...,corr len -1; the calculation formula of the fourth correlation value is as follows:

[0112]

[0113] Among them, corr_try_tmp2 is the fourth correlation value.

[0114] yf1(n),n=0,1,...,corr len -1 is the second output result.

[0115] N flt The filter gets N fltThe fourth correlation value is found from corr_try_tmp2, that is, corr_try2 is used as the fourth maximum correlation value. The corresponding fourth index number (the number of each filter group in the second filtering module) is frac_try2, and the value ranges from 1 to N flt .

[0116] Finally, compare the sizes of corr_try0, corr_try1 and corr_try2: when the maximum value is corr_try0, the integer timing deviation is idx, and there is no fractional timing deviation; when the maximum value is corr_try1, the integer timing deviation is idx, and the fractional timing deviation is frac_try1; when the maximum value is corr_try2, the integer timing deviation is idx-1, and the fractional timing deviation is frac_try2.

[0117] In step A5, the cyclic prefix is removed from the second time domain data after timing offset compensation, and FFT is performed to transform the data into the frequency domain to obtain second frequency domain data, which is recorded as Y(k), k=0, 1, ..., K-1.

[0118] In step A6, further, in order to reduce the amount of calibration data, it is possible to step The compensation is calculated for each subcarrier, and the number of calibration data is reduced to an integer. The resulting calibration data is expressed as:

[0119]

[0120] in, is the calibration data;

[0121] L(k), k=0,1,...,K-1 is the first frequency domain data;

[0122] Y(k), k=0,1,...,K-1 is the second frequency domain data;

[0123] N step is the preset number of subcarriers;

[0124] K is the total number of subcarriers within the channel bandwidth.

[0125] The calibration data of the present invention is in complex form, including both amplitude and phase unevenness.

[0126] In step A7, the service data in the time domain is received normally. After removing the cyclic prefix and performing FFT transformation on each OFDM symbol, the service data in the frequency domain is formed and recorded as Y. r (k), k=0,1,...,K-1, use calibration data to perform phase and amplitude unevenness calibration, every N stepThe same calibration result is used for each subcarrier, and the calculation formula for the calibration compensation is as follows:

[0127]

[0128] Among them, Y r (k), k=0,1,...,K-1 is the service data in the frequency domain;

[0129] Y comp (k) is the service data after calibration and compensation of the service data in the frequency domain.

[0130] By fully scheduling known modulation symbols on the channel bandwidth, calibration data in the frequency domain is obtained after timing deviation estimation. This calibration data can then be used to directly perform unevenness compensation on a subcarrier basis in the frequency domain, which can better reflect the analog filter passband amplitude and phase unevenness characteristics in the frequency domain while reducing computational complexity.

[0131] The present invention also provides an analog filter unevenness compensation system for executing the aforementioned analog filter unevenness compensation method.

[0132] The above are only preferred embodiments of the present invention and do not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for compensating for unevenness of an analog filter, characterized in that: include: Step A1: fully schedule known modulation symbols on the channel bandwidth to form first frequency domain data; Step A2: converting the first frequency domain data into the time domain to obtain first time domain data; Step A3: sending the first time domain data, and obtaining time domain received data through radio frequency processing; Step A4, performing timing offset compensation on the time domain received data according to the first time domain data to obtain second time domain data; Step A5: converting the second time domain data into the frequency domain to form second frequency domain data; Step A6: obtaining calibration data of an analog filter based on the first frequency domain data and the second frequency domain data; Step A7: receiving service data in the time domain and converting it into the frequency domain, compensating the service data in the frequency domain according to the calibration data, and then outputting the compensated data.

2. The method for compensating for unevenness of an analog filter according to claim 1, wherein: The step A4 comprises: Step A41: calculating a maximum correlation value based on the first time domain data and the time domain received data as a first maximum correlation value, obtaining a first index number associated with the first maximum correlation value, and obtaining an index number preceding the first index number as a second index number; Step A42: extracting data of a preset length from the time-domain received data starting from a preset number of sampling points in advance based on the first index number to form first target data, and extracting data of the preset length starting from the preset number of sampling points in advance based on the second index number to form second target data; Step A43, calculating the correlation value of the first target data as a second correlation value, calculating the maximum correlation value of a first output result after the first target data is filtered by the first filtering module as a third maximum correlation value, and calculating the maximum correlation value of a second output result after the second target data is filtered by the second filtering module as a fourth maximum correlation value; Step A44: Determine the numerical relationship between the second correlation value, the third maximum correlation value, and the fourth maximum correlation value: If the second correlation value is the largest, execute step A45; If the third maximum correlation value is the largest, execute step A46; If the fourth maximum correlation value is the largest, execute step A47; Step A45: Using the first index number, perform integer timing offset compensation on the time domain received data; Step A46: performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a group of filters associated with the first index number and the third maximum correlation value in the first filtering module; Step A47: performing integer timing offset compensation and fractional timing offset compensation on the time domain received data using a group of filters associated with the second index number and the fourth maximum correlation value in the second filtering module; The first filter module and the second filter module both include a plurality of filter groups, and the preset number of sampling points and the preset length are associated with the number of coefficients of each filter group.

3. The analog filter unevenness compensation method according to claim 2, wherein: The step A41 includes: Step A411: record the complete known modulation symbols in the first time domain data as x(n), n=0,1,,corr len -1 is padded with 0 at the end to a length of N and then transformed into the frequency domain. The conjugate of the complex number is then taken to obtain the first frequency domain result. At the same time, n is changed from 0 to corr len +2×offset len -1 part of the time domain received data y(n), n = 0, 1,, corr len +2×offset len -1 is padded with 0 at the end to a length of N and then transformed into the frequency domain to obtain the second frequency domain result; Step A412: multiplying the first frequency domain result and the second frequency domain result to obtain a third frequency domain result; Step A413: convert the third frequency domain result to the time domain to obtain a first time domain result, and then retain the first 2×offset of the first time domain result. len +1 data as the first relevant value; Step A414: taking the first correlation value corresponding to the maximum absolute value as the first maximum correlation value, and taking the index number of the first maximum correlation value in the first correlation values as the first index number; Among them, corr len is the associated length; offset len The preset offset points.

4. The analog filter unevenness compensation method according to claim 3, wherein: In step A42, the preset number of sampling points is rounded down. The preset length is corr len +N tap ; Among them, N tap is the number of coefficients in each filter group.

5. The analog filter unevenness compensation method according to claim 4, wherein: In the step A42, the first target data extracted is y0(n), n=0,1,,corr len +N tap -1; In step A43, the second correlation value is calculated using the following formula: Here, corr_try0 is the second correlation value.

6. The method for compensating for unevenness of an analog filter according to claim 4, wherein: In step A43, the first filter module includes N flt The first target data is filtered by each filter group in the first filter module and N tap filter output to obtain the first output result; In step A43, the third correlation value of the first output result is calculated using the following formula: Wherein, corr_try_tmp1 is the third correlation value; yf0(n),n=0,1,,corr len -1 is the first output result; In step A43, the maximum value among the third correlation values is found as the third maximum correlation value.

7. The analog filter unevenness compensation method according to claim 4, wherein: In step A43, the second filter module includes N flt The second target data is filtered by each filter group in the second filter module and N tap filter output to obtain the second output result; In step A43, the fourth correlation value of the second output result is calculated using the following formula: Wherein, corr_try_tmp2 is the fourth correlation value; yf1(n),n=0,1,,corr len -1 is the second output result; In step A43, the maximum value among the fourth correlation values is found as the fourth maximum correlation value.

8. The analog filter unevenness compensation method according to claim 4, wherein: In step A6, the calculation formula of the calibration data is as follows: in, is the calibration data; L(k), k=0, 1,, K-1 is the first frequency domain data; Y(k), k=0, 1,, K-1 is the second frequency domain data; N step is the preset number of subcarriers; K is the total number of subcarriers within the channel bandwidth.

9. The analog filter unevenness compensation method according to claim 8, wherein: In step A7, the calculation formula for compensating the service data in the frequency domain according to the calibration data is as follows: Among them, Y r (k), k = 0, 1,, K-1 is the service data in the frequency domain; Y comp (k) is the business data after calibration and compensation of the business data in the frequency domain.

10. An analog filter non-flatness compensation system, characterized in that: Used to execute the analog filter unevenness compensation method as described in any one of claims 1-9.

Images