A PAPR Reduction Method in an F-OFDM System

Abstract

The present invention discloses a method for reducing PAPR in an F-OFDM system, which relates to the technical fields of electronic information and digital signal processing, and includes the following steps: S1: Divide the spectrum into M sub-bands; S2: Perform quadrature amplitude modulation on the input signal respectively and then perform subcarrier mapping to obtain M output signals; S3: Divide the time-domain data after the inverse discrete Fourier transform of the M output signals into V groups, each group containing M / V sub-band signals with a length of N; S4: Sum the sub-band signals in each group through complex addition; S5: Generate a cyclic shift vector of length K [s1, s2, …, s k , and a total of K V phase combinations are generated; S6: Superimpose the phase combinations to obtain corresponding candidate signals; S7: Calculate the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission; By generating candidate signals, the present invention reduces the PAPR value in the system and improves the PAPR suppression performance.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic information and digital signal processing, and particularly relates to a method for reducing PAPR in an F-OFDM system. Background Art

[0002] OFDM (Orthogonal Frequency Division Multiplexing) is the orthogonal frequency division multiplexing technology. Its core idea is to divide the channel into several orthogonal sub-channels, convert the high-speed data signal into parallel low-speed sub-data streams, and modulate them for transmission on each sub-channel. Since the signal bandwidth on each sub-channel is less than the relevant bandwidth of the channel, it can effectively combat frequency selective fading, and has a high frequency band utilization rate, enabling high-speed data transmission. Moreover, its modulation and demodulation processes can be based on the discrete Fourier transform, with high realizability. Currently, it is widely used in communication systems such as LTE and 802.11a in the WLAN standard.

[0003] Since OFDM is composed of the superposition of several sub-carrier signals, it will generate a large instantaneous peak amplitude at certain moments, resulting in a higher peak-to-average power ratio (PAPR) compared to single-carrier modulation systems. This requires a higher linearity for the transmitter power amplifier. If the dynamic range of the power amplifier does not meet the signal changes, the signal will be distorted after passing through the power amplifier, destroying the orthogonality between sub-carriers and affecting the system performance. Therefore, it is necessary to reduce the relatively high PAPR.

[0004] With the advent of the 5G era, along with richer application scenarios, more flexible waveform technologies are required. Since traditional OFDM systems have disadvantages such as high out-of-band leakage and the entire bandwidth can only support one waveform parameter, they can no longer meet the requirements of 5G communication. However, filtered orthogonal frequency division multiplexing (F-OFDM) based on sub-band filtering can divide the system bandwidth into several sub-bands, and each sub-band can be configured with more flexible parameters according to different application scenarios. The sub-band filter can better suppress out-of-band radiation and further improve the spectrum utilization rate. At the same time, F-OFDM inherits the advantages of OFDM and its core idea remains unchanged, enabling a smoother transition. Therefore, F-OFDM technology has rich application scenarios and high research value. Since F-OFDM is developed on the basis of OFDM, in the F-OFDM communication system, there is still the problem of the same high PAPR value as in OFDM. Thus, it is necessary to solve the problem of high PAPR value in the F-OFDM communication system. Summary of the Invention

[0005] To solve the problem of high PAPR value in the F-OFDM communication system in the prior art, a method for reducing PAPR in the F-OFDM system is proposed to reduce the PAPR value in the F-OFDM communication system.

[0006] The technical solution of the present invention is as follows:

[0007] A method for reducing PAPR in the F-OFDM system, comprising the following steps:

[0008] S1: Divide the spectrum into M sub-bands, number each sub-carrier according to each sub-band parameter, and each sub-carrier corresponds to a different sub-band; M is a positive integer greater than zero;

[0009] S2: Respectively perform quadrature amplitude modulation on the input signals of the M sub-bands, and perform sub-carrier mapping on the data after quadrature amplitude modulation according to the number of sub-carriers to obtain M output signals;

[0010] S3: Divide the time-domain data obtained by performing inverse discrete Fourier transform on the M output signals into V groups, each group contains M / V sub-band signals, and the length of the sub-band signals is N; V is a positive integer greater than zero, and V <;

[0011] S4: Sum the sub-band signals in each group through V complex additions of length N;

[0012] S5: Use a sub-optimal established time standard setting method to generate a cyclic shift vector [s1, s2,..., s k ; where the vector element values represent the number of bits of cyclic shift, and a total of K V phase combinations are generated; K is a positive integer greater than zero, and K < V;

[0013] S6: Superimpose the generated K V phase combinations to obtain corresponding candidate signals;

[0014] S7: Calculate the PAPR values of all candidate signals respectively, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

[0015] Preferably, for S5, the sub-optimal established time standard setting method is as follows:

[0016] Multiply the summed sub-band signals by the cyclic shift vector to change their phases, thereby reducing the probability of high PAPR occurrence; the length K of the cyclic shift vector is set to 2.

[0017] Preferably, for S6, in order to reduce the amount of computation, a recursive method is used to generate candidate signals;

[0018] The method for generating candidate signals in a recursive manner is as follows:

[0019] For the signal x of the v-th group v , denote the signal after circularly shifting it by k bits as

[0020]

[0021] where k represents the number of bits of circular shift.

[0022] Superimpose the circularly shifted grouped signals to obtain candidate signals:

[0023]

[0024] Utilize the feature that there is only one different group between adjacent superimposed output signals to recursively generate all candidate signals.

[0025] Furthermore, define a superimposed candidate signal If there is only group v different between candidate signal x2 and x1, then calculate according to the following formula:

[0026]

[0027] Change the group in formula (3) from 1 to V, and generate all candidate signals in this recursive manner.

[0028] Furthermore, the calculation formula of the PAPR value is as follows:

[0029]

[0030] where E(|x[n]| 2 ) represents the average power of the time-domain signal of each sub-band.

[0031] Calculate the PAPR values of all candidate signals, compare the PAPR values of all candidate signals, and select the candidate signal with the minimum PAPR value for transmission.

[0032] Furthermore, the PAPR value is characterized by the complementary cumulative distribution function, and the expression is as follows:

[0033] CCDF(PAPR0) = Pr(PAPR > PAPR0)

[0034] The probability that the PAPR value is greater than the preset value PAPR0, CCDF represents the complementary cumulative distribution function, PAPR0 is the preset average peak power ratio, and Pr is the probability.

[0035] Preferably, M takes the value of 8, V takes the value of 4, and K takes the value of 2.

[0036] Further, the time-domain data after the inverse discrete Fourier transform of the processed data of the 8 sub-bands is divided into 4 groups, each group containing 2 sub-band signals, and the grouped time-domain data is expressed as:

[0037] (subband1, subband2), (subband3, subband4), …, (subband7, subband8)

[0038] where subband1 to subband8 represent sub-band 1 to sub-band 8; the data of two adjacent sub-bands are added to obtain 4 groups x1, x2, x3, x4.

[0039] Furthermore, the cyclic shift value is taken from [0, 1], that is, the grouped time-domain data is cyclically shifted by 0 bits or 1 bit; a total of 2 4 phase combinations are generated; the candidate signals obtained thereby are expressed as:

[0040]

[0041] In the formula,

[0042] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method are implemented.

[0043] Advantages of the present invention:

[0044] 1. Under the same conditions, it can generate more candidate signals than the traditional partial transmit sequence algorithm, achieving better PAPR suppression performance.

[0045] 2. The data grouping in the traditional partial transmit sequence algorithm is replaced by the sub-bands divided in the F-OFDM system, eliminating the grouping operation.

[0046] 3. The cyclic shift is used to replace the phase rotation factor of the traditional method, greatly reducing the number of complex multiplication calculations, reducing the system operation complexity, and adapting to systems with more sub-bands.

[0047] 4. There is no need to transmit sideband information, improving the transmission efficiency of the system.

[0048] 5. The candidate signals are generated in a recursive manner, reducing the number of complex addition operations. Description of the Drawings

[0049] Figure 1 It is a principle block diagram of a PAPR reduction method in the F-OFDM system described above. Detailed Embodiments

[0050] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] Embodiment 1

[0052] In this embodiment, as Figure 1 shown, a method for reducing PAPR in an F-OFDM system includes the following steps:

[0053] S1: Divide the spectrum into M subbands, number each subcarrier according to the parameters of each subband, and each subcarrier corresponds to a different subband; M is a positive integer greater than zero;

[0054] S2: Perform quadrature amplitude modulation on the input signals of the M subbands respectively, and perform subcarrier mapping on the data after quadrature amplitude modulation according to the subcarrier numbers to obtain M output signals;

[0055] S3: Divide the time-domain data obtained by performing inverse discrete Fourier transform on the M output signals into V groups, each group contains M / V subband signals, and the length of the subband signals is N; V is a positive integer greater than zero, and V <;

[0056] S4: Sum the subband signals in each group through V complex additions of length N;

[0057] S5: Use a suboptimal established time standard setting method to generate a cyclic shift vector [s1, s2,..., s k ; where the vector element values represent the number of bits of cyclic shift, and a total of K V phase combinations are generated; K is a positive integer greater than zero, and K < V;

[0058] S6: Superimpose the K V phase combinations generated to obtain corresponding candidate signals;

[0059] S7: Calculate the PAPR values of all candidate signals respectively, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

[0060] In this embodiment, F-OFDM can set the subcarrier spacing and modulation order of each subband separately according to the requirements in different scenarios. In order to ensure the same sampling rate of each subband, FFTs with different numbers of points need to be used. Therefore, it is necessary to uniformly number the subcarriers in the entire frequency domain, and at the same time, calculate the subcarrier coding used for each subband through the set guard interval to avoid possible crosstalk between different subbands.

[0061] The method of quadrature amplitude modulation is as follows:

[0062] Each sub - band maps the input binary data signal to points on the complex plane according to the set modulation order. For example, 16QAM means the modulation order is 16. The specific process is to divide the input binary data into groups of four each, and map them to 16 points (0000, 0001, …, 1111) on the complex plane according to the mapping rule (usually Gray code), and then multiply by the normalization factor (for 16QAM it is ) to achieve modulation. After modulation, the binary signal becomes a complex signal.

[0063] The method of sub - carrier mapping is as follows:

[0064] The complex signal after quadrature amplitude modulation needs to be modulated by OFDM, that is, transformed by IFFT into a complex signal in the time domain. In order to prevent interference in the frequency domain, it is necessary to uniformly number the sub - carriers in the entire frequency domain. For example, in an F - OFDM system, if the system is set to two sub - bands, the number of valid sub - carriers in sub - band 1 is M1, the sub - carrier spacing is f1, the number of guard sub - carriers is N1, the number of valid sub - carriers in sub - band 2 is M2, the sub - carrier spacing is f2, and the number of guard sub - carriers is N2. If the valid sub - carrier numbers in sub - band 1 are [M 1min , M 1max , then the mapped numbers of valid sub - carriers in sub - band 2 should be Numbering in this way can effectively avoid interference between sub - carrier frequencies of different sub - bands.

[0065] In this embodiment, the established time standard setting method, that is, the partial transmit sequence method, is a type of method to reduce PAPR. Its basic idea is to block the time - domain signal after IFFT and then multiply by a pre - set cyclic shift vector to change its phase, thereby reducing the probability of high PAPR occurrence. The sub - optimal established time standard setting method is the established time standard setting method with the cyclic shift vector of 2, which is to exchange a certain performance loss for a great reduction in implementation complexity. In the present invention, the length of the cyclic vector is set to 2, and more specifically, it is set to [0, 1], that is, cyclic shift by 0 or 1 bit, and the cyclic shift method corresponds to formula (1).

[0066] In this embodiment, for S5, the sub - optimal established time standard setting method is as follows:

[0067] Multiply the summed sub - band signal by the cyclic shift vector to change its phase, thereby reducing the probability of high PAPR occurrence; the length K of the cyclic shift vector is set to 2.

[0068] In this embodiment, for S6, in order to reduce the amount of computation, a recursive method is used to generate candidate signals;

[0069] The method for generating candidate signals in a recursive manner is as follows:

[0070] For the signal x of the v-th group v , denote the signal after circularly shifting it by k bits as

[0071]

[0072] where k represents the number of bits of circular shift; K refers to the length of the circular shift vector, that is, the set of all possible numbers of bits of circular shift. For example, if the generated circular shift vector is [0, 1], then the possible numbers of shift bits are 0 and 1. Then each group needs to perform these two possible circular shifts, and then all candidate signals are generated through permutation and combination. If there are 4 groups, then a total of 2 4 = 16 candidate signals can be combined, and the candidate signals are:

[0073] Superimpose the circularly shifted grouped signals to obtain candidate signals:

[0074]

[0075] Utilize the feature that there is only one group difference between adjacent superimposed output signals to recursively generate all candidate signals.

[0076] More specifically, define a superimposed candidate signal If there is only a difference in group v between candidate signal x2 and x1, then calculate according to the following formula:

[0077]

[0078] Change the group in formula (3) from 1 to V, and generate all candidate signals in this recursive manner.

[0079] More specifically, the calculation formula of the PAPR value is as follows:

[0080]

[0081] where E(|x[n]| 2 ) represents the average power of the time-domain signal of each sub-band.

[0082] Calculate the PAPR values of all candidate signals, compare the PAPR values of all candidate signals, and select the candidate signal with the minimum PAPR value for transmission.

[0083] More specifically, the PAPR is characterized using the complementary cumulative distribution function, and the expression is as follows:

[0084] CCDF(PAPR0) = Pr(PAPR > PAPR0)

[0085] The probability that the PAPR value is greater than a preset value PAPR0, CCDF represents the complementary cumulative distribution function, PAPR0 is the preset average peak power ratio, and PAPR0 is set according to requirements; Pr is the probability.

[0086] Embodiment 2

[0087] In this embodiment, a method for reducing PAPR in an F-OFDM system includes the following steps:

[0088] S1: Divide the spectrum into M subbands, number each subcarrier according to each subband parameter, and each subcarrier corresponds to a different subband; M is a positive integer greater than zero;

[0089] S2: Perform quadrature amplitude modulation on the input signals of the M subbands respectively, and perform subcarrier mapping on the data after quadrature amplitude modulation according to the subcarrier numbers to obtain M output signals;

[0090] S3: Divide the time-domain data obtained by performing the inverse discrete Fourier transform on the M output signals into V groups, each group contains M / V subband signals, and the length of the subband signals is N; V is a positive integer greater than zero, and V <;

[0091] S4: Sum the subband signals in each group through V complex additions of length N;

[0092] S5: Use a suboptimal established time standard setting method to generate a cyclic shift vector [s1, s2,..., s k ; where the vector element values represent the number of bits of cyclic shift, and a total of K V phase combinations are generated; K is a positive integer greater than zero, and K < V;

[0093] S6: Superimpose the generated K V phase combinations to obtain corresponding candidate signals;

[0094] S7: Calculate the PAPR values of all candidate signals respectively, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

[0095] In this embodiment, M takes the value of 8 and V takes the value of 4.

[0096] More specifically, the time-domain data after the inverse discrete Fourier transform (IFFT) of the processed data of 8 subbands is divided into 4 groups, and each group contains 2 subband signals. The time-domain data after grouping is expressed as:

[0097] (subband1,subband2),(subband3,subband4),…,(subband7,subband8)

[0098] Add the data of two adjacent subbands to obtain 4 groups x1, x2, x3, and x4.

[0099] More specifically, the cyclic shift value is taken from [0, 1], that is, the time-domain data after grouping is cyclically shifted by 0 bits or 1 bit; a total of 2 4 phase combinations are generated; the candidate signals obtained therefrom are expressed as:

[0100]

[0101] In the formula,

[0102] The calculation formula of the PAPR value is as follows:

[0103]

[0104] where E(|x[n]| 2 ) represents the average power of the time-domain signal of each subband.

[0105] Calculate the PAPR values of all candidate signals, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

[0106] The PAPR value is characterized by the complementary cumulative distribution function, and the expression is as follows:

[0107] CCDF(PAPR0) = Pr(PAPR > PAPR0)

[0108] The probability that the PAPR value is greater than the preset value PAPR0, CCDF represents the complementary cumulative distribution function, PAPR0 is the preset average peak power ratio, and PAPR0 is set according to requirements; Pr is the probability.

[0109] A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method are implemented.

[0110] Embodiment 3

[0111] In this embodiment, a PAPR reduction method in an F-OFDM system includes the following steps:

[0112] Divide the spectrum into 8 subbands, number each subcarrier according to the parameters of each subband, and each subcarrier corresponds to a different subband; perform quadrature amplitude modulation on the input signals of the 8 subbands respectively, and perform subcarrier mapping on the data after quadrature amplitude modulation according to the number of the subcarriers to obtain the processed data; divide the time-domain data after the inverse discrete Fourier transform (IFFT) of the processed data of the 8 subbands into 4 groups, each group contains 2 subband signals, and the length of the subband signal is N; after the time-domain data is grouped as follows:

[0113] (subband1, subband2), (subband3, subband4),..., (subband7, subband8)

[0114] Sum the subband signals in each group through N complex additions, add the data of two adjacent subbands to obtain 4 groups x1, x2, x3, x4; generate a cyclic shift vector [s1, s2,..., s k , the cyclic shift value takes [0, 1], and the grouped data is cyclically shifted by 0 bits or 1 bit, and a total of 2 4 optional phase combinations can be generated; superimpose the generated 2 4 phase combinations to obtain the corresponding candidate signals; the candidate signals are expressed as:

[0115] where Utilize the characteristic that there is only one group difference between adjacent superimposed output signals, and generate all candidate signals in a recursive manner; calculate the PAPR values of all candidate signals respectively; the calculation formula of the PAPR value is as follows:

[0116]

[0117] where E(|x[n]| 2 ) represents the average power of the time-domain signal of each subband.

[0118] Compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

[0119] Obviously, the above embodiments of the present invention are only examples for clearly illustrating the present invention, and are not limitations on the implementation manners of the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the claims of the present invention.

Claims

1. A method for reducing PAPR in an F-OFDM system, characterized in that, It includes the following steps: S1: Divide the spectrum into M sub-bands, number each sub-carrier according to the parameters of each sub-band, and each sub-carrier corresponds to a different sub-band; M is a positive integer greater than zero; S2: Perform quadrature amplitude modulation on the input signals of the M sub-bands respectively, and perform sub-carrier mapping on the data after quadrature amplitude modulation according to the number of sub-carriers to obtain M output signals; S3: Divide the time-domain data obtained by performing inverse discrete Fourier transform on the M output signals into V groups, each group contains M / V sub-band signals, and the length of the sub-band signals is N; V is a positive integer greater than zero, and V < M; S4: Sum the sub-band signals in each group through V complex additions with a length of N; S5: Generate a cyclic shift vector [s1, s2, …, s k of length K using a sub-optimal established time standard setting method; where the vector element values represent the number of bits of cyclic shift, and a total of K V phase combinations are generated; K is a positive integer greater than zero, and K < V; S6: For the generated K V phase combinations, perform superposition to obtain corresponding candidate signals; S7: Calculate the PAPR values of all candidate signals respectively, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission; For S5, the method for setting the sub-optimal established time standard is as follows: Multiply the summed sub-band signals by a cyclic shift vector to change their phases, thereby reducing the probability of high PAPR occurrences; the length K of the cyclic shift vector is set to 2; For S6, in order to reduce the amount of computation, a recursive method is used to generate candidate signals; The method for generating candidate signals in the recursive manner is as follows: For the signal x of the v-th packet v , denote the signal after cyclic shifting it by k bits as Where, k refers to the number of bits of cyclic shift; Superimpose the cyclically shifted grouped signals to obtain candidate signals: Utilize the characteristic that only one group is different between adjacent superimposed output signals, and recursively generate all candidate signals; Define a superimposed candidate signal If there is only a difference in packet v between candidate signals x2 and x1, calculate according to the following formula: Change the group in formula (3) from 1 to V, and generate all candidate signals in this recursive manner; The calculation formula of the PAPR value is as follows: where E(|x[n]| 2 ) represents the average power of the time-domain signal of each sub-band; Thus, calculate the PAPR values of all candidate signals, compare the PAPR values of all candidate signals, and select the candidate signal with the smallest PAPR value for transmission.

2. The PAPR reduction method in an F-OFDM system according to claim 1, characterized in that, The PAPR value is characterized by using the complementary cumulative distribution function, and the expression is as follows: CCDF(PAPR0) = Pr(PAPR > PAPR0) The probability that the PAPR value is greater than the pre-set value PAPR0, CCDF represents the complementary cumulative distribution function, PAPR0 is the pre-set average peak power ratio, and Pr is the probability.

3. A PAPR reduction method in an F-OFDM system according to claim 1, characterized in that, M takes the value of 8, and V takes the value of 4.

4. The PAPR reduction method in an F-OFDM system according to claim 3, characterized in that, Divide the time-domain data after inverse discrete Fourier transform of the processed data of 8 sub-bands into 4 groups, each group contains 2 sub-band signals, and the grouped time-domain data is expressed as: (subband1, subband2), (subband3, subband4), …, (subband7, subband8) where subband1~subband8 represent sub-band 1~sub-band 8; Add the data of adjacent two sub-bands to obtain 4 groups x1, x2, x3, x4.

5. A PAPR reduction method in an F-OFDM system according to claim 4, characterized in that, The cyclic shift value is taken from [0, 1], that is, the time-domain data after grouping is cyclically shifted by 0 bits or 1 bit; a total of 2 4 phase combinations are generated; The candidate signals obtained thereby are expressed as: In the formula, 6. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: When the described processor executes the described computer program, it realizes the steps of the method according to any one of claims 1~5.

Images