CN112804180A - Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing - Google Patents

Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing Download PDF

Info

Publication number
CN112804180A
CN112804180A CN202110016285.8A CN202110016285A CN112804180A CN 112804180 A CN112804180 A CN 112804180A CN 202110016285 A CN202110016285 A CN 202110016285A CN 112804180 A CN112804180 A CN 112804180A
Authority
CN
China
Prior art keywords
signal
symbol
module
oqam
estimation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110016285.8A
Other languages
Chinese (zh)
Inventor
袁晓军
欧志豪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN202110016285.8A priority Critical patent/CN112804180A/en
Publication of CN112804180A publication Critical patent/CN112804180A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2697—Multicarrier modulation systems in combination with other modulation techniques
    • H04L27/2698—Multicarrier modulation systems in combination with other modulation techniques double density OFDM/OQAM system, e.g. OFDM/OQAM-IOTA system
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2626—Arrangements specific to the transmitter only
    • H04L27/2627—Modulators
    • H04L27/264—Pulse-shaped multi-carrier, i.e. not using rectangular window
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2647—Arrangements specific to the receiver only
    • H04L27/2655—Synchronisation arrangements
    • H04L27/2689—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation
    • H04L27/2691—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation involving interference determination or cancellation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2647—Arrangements specific to the receiver only
    • H04L27/2655—Synchronisation arrangements
    • H04L27/2689—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation
    • H04L27/2695—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation with channel estimation, e.g. determination of delay spread, derivative or peak tracking

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Digital Transmission Methods That Use Modulated Carrier Waves (AREA)

Abstract

The invention belongs to the technical field of information and communication, and particularly relates to an OQAM/FBMC system signal receiving and transmitting method based on amplitude limiting of compressed sensing. The invention has sparsity in time domain according to the distortion caused by amplitude limiting processing on OQAM/FBMC signals, namely, only a small part of the signal distortion in the time domain is a non-zero value, and the distortion can be well estimated and compensated by utilizing a Compressive Sensing (CS) method, thereby greatly reducing the PAPR under the condition of ensuring the error rate of a system. Estimating a truncated portion z of the signal while effectively reducing the PAPR of the signal to more accurately recover the signal; the invention carries out the Clipping and Filtering processing and uses the Turbo-CS algorithm to estimate the distorted signal in the time domain, on one hand, the PAPR of the signal is reduced by the Clipping at the transmitting end and the Filtering is used to relieve the side lobe regeneration, on the other hand, the Turbo-CS algorithm is used at the receiving end to detect the distorted signal z to ensure the performance of the error rate of the system.

Description

Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing
Technical Field
The invention belongs to the technical field of information and communication, and particularly relates to an OQAM/FBMC system signal receiving and transmitting method based on amplitude limiting of compressed sensing.
Background
As an alternative technology to OFDM, an Offset Quadrature Amplitude Modulation based Filter Bank Multi-carrier (OQAM/FBMC) has the advantages of low out-of-band leakage of frequency spectrum, low sensitivity to synchronization deviation, and higher robustness to multipath channels, and is currently widely researched. However, like other multi-carrier modulation techniques, the OQAM/FBMC technique also has a high PAPR problem, and the PAPR problem becomes more complicated due to the overlapping of the OQAM/FBMC symbols. At present, means for reducing PAPR of an OQAM/FBMC signal mainly include a signal distortion and distortion-free reduction method, the distortion-free method needs to overcome the problem of obtaining an optimal reduction effect under a low complexity condition and simultaneously reduce overhead, and the signal distortion reduction method needs to avoid a side lobe regeneration phenomenon caused by signal distortion and reduce in-band distortion of the signal as much as possible.
A simple and effective method is to clip the signal in advance, i.e. before the carrier signal passes through the high power amplifier, etc., the time domain signal is clipped, if the peak value of the signal exceeds a given threshold value, the signal is set to a preset threshold value while the phase is unchanged, and if the peak value of the signal is not greater than the threshold value, no operation is performed.
Figure BDA0002886816250000011
Where a is a set amplitude threshold value,
Figure BDA0002886816250000012
representing the phase of the symbol. The clipping method is simple to operate and low in computational complexity, and can significantly improve the PAPR performance of the system, but belongs to distortion transformation, and can cause serious nonlinear distortion to signals. .
Disclosure of Invention
In order to solve the problems in the prior art, the distortion caused by amplitude limiting processing on the OQAM/FBMC signal has sparsity in the time domain, namely only a small part of the signal distortion in the time domain is a non-zero value, and the distortion can be well estimated and compensated by using a Compressive Sensing (CS) method, so that the PAPR is greatly reduced under the condition of ensuring the system error rate. Therefore, the signal transceiving method of the OQAM/FBMC system based on amplitude limiting of compressed sensing is provided.
The receiver of the invention consists of three modules, namely a linear estimation module A, a signal demodulation module B and an amplitude limiting estimation noise reduction module C. Each module outputs an estimation and transmits the estimation to the next module, the OQAM/FBMC symbol a is iteratively estimated by the module A and the module B, the sparse distortion z is iteratively estimated by the module A and the module C, and iteration is carried out among the modules until the algorithm is converged.
The technical scheme adopted by the invention is as follows: a signal transceiving method of an OQAM/FBMC system based on compressed sensing amplitude limiting comprises the following steps:
s1, inputting binary bit stream b [ k ]]QAM modulating to obtain a mapping symbol c with length Mm，nAs an OQAM/FBMC symbol, two real symbols with the length of M are obtained by taking a real part and an imaginary part
Figure BDA0002886816250000021
And
Figure BDA0002886816250000022
noting that the nth transmitted symbol is
Figure BDA0002886816250000023
Carrying out phase rotation, frequency domain filtering and inverse fast Fourier transform on the transmitted symbol to obtain a length LgUp-sampled signal vector s of KM (K is an overlap factor indicating the number of overlap of OQAM/FBMC symbols)n，
sn＝FHGΘnan
Wherein the filter matrix is
Figure BDA0002886816250000024
GmIs the length L on the m sub-carriergFrequency domain response, phase rotation of prototype filter
Figure BDA0002886816250000025
θm，n＝ejπ(m+n)/2ejnmπF is Lg×LgUnit DFT matrix, (.)HRepresenting a conjugate transpose.
S2, taking N OQAM/FBMC symbols as a data frame, delaying the generated 2N signal vectors by half a symbol period, namely M/2 sampling points in turn, generating a transmitting signal after superposition,
s=Wa
wherein
Figure BDA0002886816250000026
Let Wn＝FHGΘnThen the modulation matrix is
Figure BDA0002886816250000027
The internal structure is shown in figure 1, and the total length of the transmitted signal is
Figure BDA0002886816250000028
The s is clipped and filtered and then transmitted to the transmitting antenna.
And S3, the signal is transmitted by the transmitting antenna, the signal passes through a multipath channel, and the tail value of the prototype filter is close to 0, so the process is a process of circularly convolving the signal and the channel.
S4, the signal arrives at the receiving end to obtain the length LsThe observed signal of (a) is,
r＝HWa+Hz+n
wherein H is a channel response matrix, z is an interference signal generated by the Clipping and filtering operation, n is zero-mean Gaussian white noise, and the noise variance is
Figure BDA0002886816250000031
S5, the receiving end cancels the interference and estimates the symbol a, specifically: the receiver comprises a linear estimation module A, a signal demodulation module B and an amplitude limiting estimation noise reduction module C, an OQAM/FBMC symbol a is iteratively estimated through the linear estimation module A and the signal demodulation module B, and sparse distortion z is iteratively estimated through the amplitude limiting estimation noise reduction module C;
the specific method of iteration is as follows:
initializing iterative receiver parameters:
Figure BDA0002886816250000032
wherein the numerical value
Figure BDA0002886816250000033
Is transmitted from the transmitting end to the receiving end,
Figure BDA0002886816250000034
the method comprises the following steps that a mean value of z is represented, I represents an identity matrix, V represents a covariance matrix, an superscript pri is defined to represent prior information, a superscript post represents posterior information, a superscript ext represents external information, a subscript A represents a corresponding parameter input/output linear estimation module A, a subscript B represents a corresponding parameter input/output signal demodulation module B, a subscript C represents a corresponding parameter input/output amplitude limiting estimation noise reduction module C, a subscript k represents a kth symbol in a vector, and subscripts a and z of the covariance matrix V respectively represent covariance which is the symbol a and amplitude limiting distortion z; the signal processing mode of each module of the receiver is as follows:
s51, inputting the received signal into a linear estimation module A, passing the signal through an analysis filter bank before the linear estimation module A estimates a, firstly passing the received signal through a filter bank with the length of LgThe rectangular sliding window obtains a section of signal by moving M/2 sampling points, and the symbol estimation value is obtained by FFT transformation, matched filtering, phase rotation and real part operation
Figure BDA0002886816250000035
Figure BDA0002886816250000036
S52, combining the clipping noise and the channel noise into demodulation noise η ═ WHHz+WHn is and have
Figure BDA0002886816250000037
Note the book
Figure BDA0002886816250000038
Taking the real part of the output of the analysis filter bank, and then carrying out single-point LMMSE estimation, wherein k is (n-1) M + M, and then the estimated value of the kth symbol is as follows:
Figure BDA0002886816250000039
the MSE matrix calculation formula is:
Figure BDA0002886816250000041
wherein HmIs the channel frequency domain response;
the linear estimation module A outputs an estimated symbol
Figure BDA0002886816250000042
And
Figure BDA0002886816250000043
s53, according to
Figure BDA0002886816250000044
And
Figure BDA0002886816250000045
computing extrinsic information
Figure BDA0002886816250000046
And
Figure BDA0002886816250000047
Figure BDA0002886816250000048
Figure BDA0002886816250000049
inputting the external information into a signal demodulation module B;
s54, in the signal demodulation module B, firstly reconstructing QAM symbol from OQAM symbol, and making
Figure BDA00028868162500000410
The reconstructed symbol is recorded as
Figure BDA00028868162500000411
Then, the estimated value of the symbol s is optimized through soft demodulation, and each constellation point is assumed to contain q bits, so that the QAM modulation order is 2qThe symbol corresponding to the constellation point is marked as cl，l＝1，2，...，2qThe log-likelihood ratio of the jth bit of the ith symbol output by the soft demodulator is recorded as lambdai，jI 1, 2., NM, j 1, 2.,. q, then the probability that the ith symbol corresponds to the ith constellation point is:
Figure BDA00028868162500000412
wherein b isi，jDenotes the ith symbol
Figure BDA00028868162500000413
J-th bit of cl，jRepresenting the ith constellation point
Figure BDA00028868162500000414
Has a probability of Pr (b) of 1i，j＝1|λi，j)＝exp(λi，j)/(1+exp(λi，j))；
S55, corresponding to the OQAM symbol structure, separately calculating the real part and the imaginary part to obtain the symbol posterior information in the nth period as follows:
Figure BDA00028868162500000415
Figure BDA00028868162500000416
wherein 2(N-1) M +1 ≤ k ≤ 2nM, N ═ 1, 2.., N;
after passing through the signal demodulation module B, posterior information is obtained
Figure BDA0002886816250000051
And
Figure BDA0002886816250000052
will posterior information
Figure BDA0002886816250000053
And
Figure BDA0002886816250000054
the input linear estimation module A updates parameters:
Figure BDA0002886816250000055
s56, in the linear estimation module A, LMMSE estimation is carried out on the amplitude limiting distortion z:
Figure BDA0002886816250000056
the MSE matrix calculation formula is as follows,
Figure BDA0002886816250000057
wherein
Figure BDA0002886816250000058
Is the channel white Gaussian noise variance, and r is the receiving end signal.
Linear estimation Module A output
Figure BDA0002886816250000059
And
Figure BDA00028868162500000510
s57, according to
Figure BDA00028868162500000511
And
Figure BDA00028868162500000512
computing extrinsic information
Figure BDA00028868162500000513
And
Figure BDA00028868162500000514
Figure BDA00028868162500000515
Figure BDA00028868162500000516
will be foreign information
Figure BDA00028868162500000517
And
Figure BDA00028868162500000518
inputting an amplitude limiting estimation noise reduction module C;
s58, estimating in an amplitude-limiting estimation noise reduction module C based on amplitude-limiting distortion sparsity, and assuming that z obeys Gaussian Bernoulli distribution, namely
Figure BDA00028868162500000519
Lambda is the clipping distortion sparsity and beta is the gaussian distribution variance.
Order to
Figure BDA00028868162500000520
The MMSE estimation calculation formula is as follows,
Figure BDA00028868162500000521
wherein
Figure BDA0002886816250000061
Figure BDA0002886816250000062
Figure BDA0002886816250000063
The posterior variance is as follows,
Figure BDA0002886816250000064
then taking the mean to approximate the posterior variance, i.e.
Figure BDA0002886816250000065
Clipping estimation noise reduction module C output
Figure BDA0002886816250000066
And
Figure BDA0002886816250000067
s59, according to
Figure BDA0002886816250000068
And
Figure BDA0002886816250000069
computing extrinsic information
Figure BDA00028868162500000610
And
Figure BDA00028868162500000611
Figure BDA00028868162500000612
Figure BDA00028868162500000613
will be foreign information
Figure BDA00028868162500000614
And
Figure BDA00028868162500000615
the input linearity estimation module A is based on parameters, i.e.
Figure BDA00028868162500000616
And S510, if the algorithm is converged, ending, otherwise, returning to the step S52.
The invention has the beneficial effects that: estimating a truncated portion z of the signal while effectively reducing the PAPR of the signal to more accurately recover the signal; the invention carries out the Clipping and Filtering processing and uses the Turbo-CS algorithm to estimate the distorted signal in the time domain, on one hand, the PAPR of the signal is reduced by the Clipping at the transmitting end and the Filtering is used to relieve the side lobe regeneration, on the other hand, the Turbo-CS algorithm is used at the receiving end to detect the distorted signal z to ensure the performance of the error rate of the system.
Drawings
Fig. 1 is a diagram of a transmit-end modulation matrix structure;
FIG. 2 is a flow chart of a method of the present invention;
FIG. 3 is a graph showing the influence of the Clipping and Filtering performed by the transmitter on the PAPR;
FIG. 4 is a simulation curve of bit error rates of different processing modes in the OQAM/FBMC system.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
As shown in fig. 2, a specific signal processing flow of the present invention includes repeated Clipping and Filtering at the transmitting end, introduction of channels, and Turbo-CS iterative receiver at the receiving end.
Setting the carrier number M used at the transmitting end to be 72, the number N of OQAM/FBMC symbols in a data frame to be 10, and the modulation mode is16-order QAM modulation with a symbol overlap factor K of 4, using a PHYDYAS project prototype filter with a filter length of 288 samples, defined
Figure BDA0002886816250000071
Wherein A is a threshold value, the amplitude limiting rate is set to be CR equal to 2dB, 1-time Clipping and Filtering processing is carried out at a transmitting end, and a simulation channel is an AWGN channel.
According to the setting of the parameters, the specific steps are as follows:
s1, inputting binary bit stream b [ k ]]QAM modulating to obtain a mapping symbol c with length of 72m，nAs an OQAM/FBMC symbol, two real symbols with the length of 72 are obtained by taking a real part and an imaginary part
Figure BDA0002886816250000072
And
Figure BDA0002886816250000073
noting that the nth transmitted symbol is
Figure BDA0002886816250000074
The transmitted symbol is subjected to phase rotation, frequency domain filtering and inverse fast fourier transform to obtain an up-sampled signal vector s of length 288n，
Sn=FHGΘnan
Wherein the filter matrix is
Figure BDA0002886816250000075
GmFor the long 288 prototype filter frequency domain response on the m sub-carrier, phase rotates
Figure BDA0002886816250000076
θm，n=ejπ(m-n)/2ejnmπF is Lg×LgUnit DFT matrix, (.)HRepresenting a conjugate transpose.
S2, taking 10 OQAM/FBMC symbols as a data frame, delaying the generated 20 signal vectors by half a symbol period in turn, namely 36 sampling points, generating a transmitting signal after superposition,
s＝Wa
wherein
Figure BDA0002886816250000077
Note Wn＝FHGΘnW is the modulation matrix, and L is the total length of the transmitted signals972. The s is subjected to Clipping and Filtering once and then transmitted to a transmitting antenna.
S3, the signal passes through AWGN channel, and the channel response matrix is the identity matrix I.
S4, the signal arrives at the receiving end to obtain the length LsThe observed signal of (a) is,
r＝HWa+Hz+n
where H ═ I is the channel response matrix, z is the interference signal generated by the Clipping and filtering operation, it is necessary to estimate and cancel the interference at the receiving end and estimate the gaussian white noise with symbol a, n being zero mean.
S5, the receiving end cancels the interference and estimates the symbol a, specifically:
initializing iterative receiver parameters:
Figure BDA0002886816250000081
wherein the numerical value
Figure BDA0002886816250000082
Is transmitted from the transmitting end to the receiving end,
Figure BDA0002886816250000083
the mean value of z is represented, I represents the identity matrix, and we use the abbreviation "pri" with the superscript "prior" to represent prior information, the abbreviation "post" with the superscript "posterior" to represent posterior information, and the abbreviation "ext" with the superscript "externic" to represent extrinsic information.
S51, before module A estimates a, the signal is passed through an analysis filter bank, and the received signal is first passed through a filter with length Lg288 rectangular sliding window, taking a segment of signal every 36 sampling points, FFT transforming, matched filtering, phase rotating and real-taking operationTo obtain a symbol estimate
Figure BDA0002886816250000084
Figure BDA0002886816250000085
The matched filter coefficients are the conjugate of the frequency domain response function of the prototype filter, so the matrix form is
Figure BDA0002886816250000086
S52, combining the clipping noise and the channel noise into demodulation noise η ═ WHHz+WHn is and have
Figure BDA0002886816250000087
Taking the real part of the output of the analysis filter bank, and then carrying out single-point LMMSE estimation, wherein k is 72(n-1) + m, the estimated value of the kth symbol is,
Figure BDA0002886816250000091
the MSE matrix is calculated by the formula,
Figure BDA0002886816250000092
wherein HmIs the channel frequency domain response.
S53, the extrinsic information calculation formula is as follows,
Figure BDA0002886816250000093
Figure BDA0002886816250000094
inputting the external information into a module B, wherein
Figure BDA0002886816250000095
S54, reconstructing QAM symbol from OQAM symbol in module B
Figure BDA0002886816250000096
The reconstructed symbol is recorded as
Figure BDA0002886816250000097
The estimate of the symbol s is then optimized by soft demodulation. Assuming that each constellation point contains 4 bits, so the QAM modulation order is 16, and the symbol corresponding to the constellation point is marked as c l1, 2.. 16. the log-likelihood ratio of the jth bit of the ith symbol output by the soft demodulator is denoted as λi，jI 1, 2, 720, j 1, 2, 4, the probability that the ith symbol corresponds to the ith constellation point is,
Figure BDA0002886816250000098
wherein b isi，jDenotes the ith symbol
Figure BDA0002886816250000099
J-th bit of cl，jRepresenting the ith constellation point
Figure BDA00028868162500000910
Has a probability of Pr (b) of 1i，j＝1|λi，j)＝exp(λi，j)/(1+exp(λi，j))。
S55, corresponding to OQAM symbol structure, calculating the real part and imaginary part separately to obtain the symbol posterior information in the nth period,
Figure BDA00028868162500000911
Figure BDA00028868162500000912
wherein 144(n-1) +1 ≦ k ≦ 144n, n ═ 1, 2
Using the posterior information of module B as input to module A, i.e.
Figure BDA0002886816250000101
S56, LMMSE estimation is carried out on the amplitude limiting distortion z in the module A, the calculation formula of the estimation value is as follows,
Figure BDA0002886816250000102
the MSE matrix calculation formula is as follows,
Figure BDA00028868162500001011
s57, the extrinsic information calculation formula is as follows,
Figure BDA0002886816250000103
Figure BDA0002886816250000104
the external information is transmitted to an amplitude limiting estimation noise reduction module C, and the order is
Figure BDA0002886816250000105
S58, estimating based on the clipping distortion sparsity in a module C, wherein the MMSE estimation calculation formula is as follows,
Figure BDA0002886816250000106
wherein
Figure BDA0002886816250000107
Figure BDA0002886816250000108
Figure BDA0002886816250000109
The posterior variance is as follows,
Figure BDA00028868162500001010
then taking the mean to approximate the posterior variance, i.e.
Figure BDA0002886816250000111
S59, an external information calculation formula,
Figure BDA0002886816250000112
Figure BDA0002886816250000113
passing the extrinsic information to module A, i.e.
Figure BDA0002886816250000114
And S510, if the algorithm is converged, ending, otherwise, returning to the step S52.
FIG. 3 is a graph showing the effect of Clipping and Filtering on the PAPR of a signal at a transmitter, where the abscissa indicates that the symbol power of dB-converted transmission is greater than the average power, and the ordinate showsThe ratio of the symbols is shown. It can be seen that 10 is taken from CCDF (complementary Current Distribution function)-5When the method is used, the PAPR is reduced by about 5dB after one-time Clipping and Filtering, and the obvious inhibition effect is achieved.
FIG. 4 shows the bit error rate performance of different systems under OQAM/FBMC system, where the bit error rate of the system is limited to BER 10-4And in time, the performance gain of about 3.6dB is obtained after one-time iterative compensation, and the performance gain of multiple-time iterative compensation is smaller. In addition, with the improvement of the signal to noise ratio, the difference of the performance curve of the algorithm of the invention and the performance curve of the algorithm without the Clipping operation is very small, and the PAPR is reduced while the detection performance is still very good.

Claims (1)

1. A signal transceiving method of an OQAM/FBMC system based on compressed sensing amplitude limiting is characterized by comprising the following steps:
s1, inputting binary bit stream b [ k ]]Obtaining a mapping symbol c with the length of M through QAM modulationm，nAs an OQAM/FBMC symbol, two real symbols with the length of M are obtained by taking a real part and an imaginary part
Figure FDA0002886816240000011
And
Figure FDA0002886816240000012
noting that the nth transmitted symbol is
Figure FDA0002886816240000013
Carrying out phase rotation, frequency domain filtering and inverse fast Fourier transform on the transmitted symbol to obtain a length LgKM-up-sampled signal vector snAnd K is an overlapping factor and represents the overlapping number of OQAM/FBMC symbols:
sn＝FHGΘnan
wherein the filter matrix is
Figure FDA0002886816240000014
GmIs the m-th subcarrierLength L of wavelengthgThe prototype filter frequency domain response, phase rotation
Figure FDA0002886816240000015
θm，n＝ejπ(m+n)/2ejnmπF is Lg×LgUnit DFT matrix, (.)HRepresents a conjugate transpose;
s2, taking N OQAM/FBMC symbols as a data frame, delaying the generated 2N signal vectors by half a symbol period in sequence, namely M/2 sampling points, and generating a transmission signal after superposition:
s＝Wa
wherein
Figure FDA0002886816240000016
Modulation matrix Wn＝FHGΘnTotal length of transmitted signal is
Figure FDA0002886816240000017
The s is subjected to Clipping and Filtering processing and then is transmitted to a transmitting antenna;
s3, the signal is transmitted by the transmitting antenna, the signal passes through the multipath channel, because the tail value of the prototype filter is close to 0, the process is the process of the cyclic convolution of the signal and the channel;
s4, receiving the signal by the receiving end to obtain the length LsThe observed signals of (a) are:
r＝HWa+Hz+n
wherein H is a channel response matrix, z is an interference signal generated by the Clipping and filtering operation, n is zero-mean Gaussian white noise, and the noise variance is
Figure FDA0002886816240000018
S5, the receiving end cancels the interference and estimates the symbol a, specifically: the receiver comprises a linear estimation module A, a signal demodulation module B and an amplitude limiting estimation noise reduction module C, an OQAM/FBMC symbol a is iteratively estimated through the linear estimation module A and the signal demodulation module B, and sparse distortion z is iteratively estimated through the amplitude limiting estimation noise reduction module C;
the specific method of iteration is as follows:
initializing iterative receiver parameters:
Figure FDA0002886816240000021
wherein the numerical value
Figure FDA0002886816240000022
Is transmitted from the transmitting end to the receiving end,
Figure FDA0002886816240000023
the method comprises the following steps that a mean value of z is represented, I represents an identity matrix, V represents a covariance matrix, an superscript pri is defined to represent prior information, a superscript post represents posterior information, a superscript ext represents external information, a subscript A represents a corresponding parameter input/output linear estimation module A, a subscript B represents a corresponding parameter input/output signal demodulation module B, a subscript C represents a corresponding parameter input/output amplitude limiting estimation noise reduction module C, a subscript k represents a kth symbol in a vector, and subscripts a and z of the covariance matrix V respectively represent covariance which is the symbol a and amplitude limiting distortion z; the signal processing mode of each module of the receiver is as follows:
s51, inputting the received signal into a linear estimation module A, passing the signal through an analysis filter bank before the linear estimation module A estimates a, firstly passing the received signal through a filter bank with the length of LgThe rectangular sliding window obtains a section of signal by moving M/2 sampling points, and the symbol estimation value is obtained by FFT transformation, matched filtering, phase rotation and real part operation
Figure FDA0002886816240000024
Figure FDA0002886816240000025
S52, combining the clipping noise and the channel noise into demodulation noise η ═ WHHz+WHn is and have
Figure FDA0002886816240000026
Note the book
Figure FDA0002886816240000027
Taking the real part of the output of the analysis filter bank, and then carrying out single-point LMMSE estimation, wherein k is (n-1) M + M, and then the estimated value of the kth symbol is as follows:
Figure FDA0002886816240000028
the MSE matrix calculation formula is:
Figure FDA0002886816240000029
wherein HmIs the channel frequency domain response;
the linear estimation module A outputs an estimated symbol
Figure FDA00028868162400000210
And
Figure FDA00028868162400000211
s53, according to
Figure FDA00028868162400000212
And
Figure FDA00028868162400000213
computing extrinsic information
Figure FDA00028868162400000214
And
Figure FDA00028868162400000215
Figure FDA00028868162400000216
Figure FDA00028868162400000217
inputting the external information into a signal demodulation module B;
s54, in the signal demodulation module B, firstly reconstructing QAM symbol from OQAM symbol, and making
Figure FDA0002886816240000031
The reconstructed symbol is recorded as
Figure FDA0002886816240000032
NM, then optimizing the estimate of the symbol s by soft demodulation, assuming that each constellation point contains q bits, so the QAM modulation order is 2qThe symbol corresponding to the constellation point is marked as cl，l＝1，2，...，2qThe log-likelihood ratio of the jth bit of the ith symbol output by the soft demodulator is recorded as lambdai，jI 1, 2., NM, j 1, 2.,. q, then the probability that the ith symbol corresponds to the ith constellation point is:
Figure FDA0002886816240000033
wherein b isi，jDenotes the ith symbol
Figure FDA0002886816240000034
J-th bit of cl，jRepresenting the ith constellation point
Figure FDA0002886816240000035
Has a probability of Pr (b) of 1i，j＝1|λi，j)＝exp(λi，j)/(1+exp(λi，j))；
S55, corresponding to the OQAM symbol structure, separately calculating the real part and the imaginary part to obtain the symbol posterior information in the nth period as follows:
Figure FDA0002886816240000036
Figure FDA0002886816240000037
wherein 2(N-1) M +1 ≤ k ≤ 2nM, N ═ 1, 2.., N;
after passing through the signal demodulation module B, posterior information is obtained
Figure FDA0002886816240000038
And
Figure FDA0002886816240000039
will posterior information
Figure FDA00028868162400000310
And
Figure FDA00028868162400000311
the input linear estimation module A updates parameters:
Figure FDA00028868162400000312
s56, in the linear estimation module A, LMMSE estimation is carried out on the amplitude limiting distortion z:
Figure FDA00028868162400000313
the MSE matrix calculation formula is as follows,
Figure FDA00028868162400000314
wherein
Figure FDA00028868162400000315
Is a channel Gaussian white noise variance, and r is a receiving end signal;
linear estimation Module A output
Figure FDA00028868162400000316
And
Figure FDA00028868162400000317
s57, according to
Figure FDA0002886816240000041
And
Figure FDA0002886816240000042
computing extrinsic information
Figure FDA0002886816240000043
And
Figure FDA0002886816240000044
Figure FDA0002886816240000045
Figure FDA0002886816240000046
will be foreign information
Figure FDA0002886816240000047
And
Figure FDA0002886816240000048
inputting an amplitude limiting estimation noise reduction module C;
s58, estimating in an amplitude-limiting estimation noise reduction module C based on amplitude-limiting distortion sparsity, and assuming that z obeys Gaussian Bernoulli distribution, namely
Figure FDA0002886816240000049
Lambda is amplitude limiting distortion sparsity, and beta is Gaussian distribution variance;
order to
Figure FDA00028868162400000410
The MMSE estimation calculation formula is as follows,
Figure FDA00028868162400000411
wherein beta is
Figure FDA00028868162400000412
Figure FDA00028868162400000413
Figure FDA00028868162400000414
The posterior variance is as follows,
Figure FDA00028868162400000415
then taking the mean to approximate the posterior variance, i.e.
Figure FDA00028868162400000416
Clipping estimation noise reduction module C output
Figure FDA00028868162400000417
And
Figure FDA00028868162400000418
s59, according to
Figure FDA00028868162400000419
And
Figure FDA00028868162400000420
computing extrinsic information
Figure FDA00028868162400000421
And
Figure FDA00028868162400000422
Figure FDA0002886816240000051
Figure FDA0002886816240000052
will be foreign information
Figure FDA0002886816240000053
And
Figure FDA0002886816240000054
the input linear estimation module A updates the parameters, i.e.
Figure FDA0002886816240000055
And S510, if the algorithm is converged, ending, otherwise, returning to the step S52.
CN202110016285.8A 2021-01-07 2021-01-07 Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing Pending CN112804180A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110016285.8A CN112804180A (en) 2021-01-07 2021-01-07 Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110016285.8A CN112804180A (en) 2021-01-07 2021-01-07 Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing

Publications (1)

Publication Number Publication Date
CN112804180A true CN112804180A (en) 2021-05-14

Family

ID=75808774

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110016285.8A Pending CN112804180A (en) 2021-01-07 2021-01-07 Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing

Country Status (1)

Country Link
CN (1) CN112804180A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114285524A (en) * 2021-12-24 2022-04-05 北京中科晶上科技股份有限公司 LLR vector quantization calculation method, device and communication system

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104823402A (en) * 2012-11-29 2015-08-05 交互数字专利控股公司 Reduction of spectral leakage in OFDM system
CN105635022A (en) * 2015-12-29 2016-06-01 中国科学院上海微系统与信息技术研究所 Offset orthogonal multicarrier baseband system
CN106506430A (en) * 2016-11-30 2017-03-15 黑龙江科技大学 A kind of new algorithm of the compensation peak-to-average force ratio non-linear distortion based on compressed sensing technology
CN106506428A (en) * 2016-11-21 2017-03-15 重庆邮电大学 Reduce the PTS method of the low computation complexity of UFMC systems PAPR
CN106713211A (en) * 2017-01-19 2017-05-24 桂林电子科技大学 OFDM/OQAM hybrid PAPR suppression system and method
CN107276726A (en) * 2017-05-16 2017-10-20 西安交通大学 A kind of Massive MIMO FBMC beam space time coding downlink transmission methods
CN107528806A (en) * 2017-05-03 2017-12-29 重庆邮电大学 A kind of SACI TR algorithms of reduction FBMC OQAM PAR peak to average ratios
US20180254754A1 (en) * 2017-03-01 2018-09-06 Paulo Carvalho Apparatus for quantized linear amplification with nonlinear amplifiers
WO2019027341A1 (en) * 2017-08-04 2019-02-07 Autonomous Non-Profit Organization For Higher Education "Skolkovo Institute Of Science And Technology" Selective tone reservation for papr reduction in wireless communication systems
CN110324271A (en) * 2019-07-22 2019-10-11 电子科技大学 A kind of F-OFDM system transceiver design method of compressed sensing based clipping
US10680870B2 (en) * 2016-11-09 2020-06-09 Huawei Technologies Co., Ltd. Synthesis of near-constant modulus waveform for high frequency transmission

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104823402A (en) * 2012-11-29 2015-08-05 交互数字专利控股公司 Reduction of spectral leakage in OFDM system
CN105635022A (en) * 2015-12-29 2016-06-01 中国科学院上海微系统与信息技术研究所 Offset orthogonal multicarrier baseband system
US10680870B2 (en) * 2016-11-09 2020-06-09 Huawei Technologies Co., Ltd. Synthesis of near-constant modulus waveform for high frequency transmission
CN106506428A (en) * 2016-11-21 2017-03-15 重庆邮电大学 Reduce the PTS method of the low computation complexity of UFMC systems PAPR
CN106506430A (en) * 2016-11-30 2017-03-15 黑龙江科技大学 A kind of new algorithm of the compensation peak-to-average force ratio non-linear distortion based on compressed sensing technology
CN106713211A (en) * 2017-01-19 2017-05-24 桂林电子科技大学 OFDM/OQAM hybrid PAPR suppression system and method
US20180254754A1 (en) * 2017-03-01 2018-09-06 Paulo Carvalho Apparatus for quantized linear amplification with nonlinear amplifiers
CN107528806A (en) * 2017-05-03 2017-12-29 重庆邮电大学 A kind of SACI TR algorithms of reduction FBMC OQAM PAR peak to average ratios
CN107276726A (en) * 2017-05-16 2017-10-20 西安交通大学 A kind of Massive MIMO FBMC beam space time coding downlink transmission methods
WO2019027341A1 (en) * 2017-08-04 2019-02-07 Autonomous Non-Profit Organization For Higher Education "Skolkovo Institute Of Science And Technology" Selective tone reservation for papr reduction in wireless communication systems
CN110324271A (en) * 2019-07-22 2019-10-11 电子科技大学 A kind of F-OFDM system transceiver design method of compressed sensing based clipping

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HANEN BOUHADDA等: "Iterative receiver cancellation of nonlinear power amplifier distortion in FBMC/OQAM system", 《2015 IEEE 11TH INTERNATIONAL CONFERENCE ON WIRELESS AND MOBILE COMPUTING, NETWORKING AND COMMUNICATIONS (WIMOB)》 *
WENJIN JIANG等: "Sparsity-Learning-Based Iterative Compensation for Filtered-OFDM With Clipping", 《IEEE COMMUNICATIONS LETTERS》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114285524A (en) * 2021-12-24 2022-04-05 北京中科晶上科技股份有限公司 LLR vector quantization calculation method, device and communication system
CN114285524B (en) * 2021-12-24 2024-01-23 北京中科晶上科技股份有限公司 LLR vector quantization calculation method, device and communication system

Similar Documents

Publication Publication Date Title
US7031251B2 (en) Clipping distortion canceller for OFDM signals
US8054900B2 (en) Symmetrical data signal processing
JP4272665B2 (en) Apparatus, method, and computer program for estimating channel of OFDM transmission system
CN106549888B (en) A kind of estimation of joint doubly selective channel and FTNS detection method based on GAMP
CN110324271B (en) Amplitude limiting F-OFDM system transceiver design method based on compressed sensing
CN112087266B (en) Time-varying broadband Doppler compensation method based on EMD-WFFT
CN112003804B (en) Multipath multivariate differential chaotic shift keying iterative receiving method
CN114285495B (en) Underwater sound OFDM communication system received signal processing method based on message transmission
Baxley et al. MAP metric for blind phase sequence detection in selected mapping
Ahmed et al. Low-complexity iterative method of equalization for single carrier with cyclic prefix in doubly selective channels
CN112804180A (en) Amplitude limiting OQAM/FBMC system signal transceiving method based on compressed sensing
Abd El-Galil et al. Efficient underwater acoustic communication with peak-to-average power ratio reduction and channel equalization
Wu Iterative compressive sensing for the cancellation of clipping noise in underwater acoustic OFDM system
Wu et al. Estimation of doubly spread underwater acoustic channel via gram-schmidt matching pursuit
Wang et al. A CFO Estimation Method Based on Preamble Symbols for GFDM
Rahmati et al. Iterative reconstruction of oversampled OFDM signals over deep fading channels
Zheng et al. A Novel Single Carrier Frequency Domain Equalization Method in Underwater Acoustic Communications
CN115865579B (en) Roll-off coefficient blind estimation method of digital modulation signal based on deep learning
Zheng et al. Deep learning-aided receiver against nonlinear distortion of HPA in OFDM systems
Anupriya et al. Channel estimation for MIMO-OFDM systems using superimposed training In VLSI
Gao et al. Clipping noise cancellation for OFDM systems based on priori aided compressed sensing
CN111327550B (en) Frequency division multiple access communication method and system with approximate constant envelope waveform
Chen et al. Frequency domain turbo equalization with iterative channel estimation for single carrier MIMO underwater acoustic communications
Bittner et al. Equalisation of MIMO-OFDM signals affected by phase noise and clipping and filtering
Mulinde DOA estimation using Constant-Envelope OFDM waveforms in RadCom fusion systems

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210514

WD01 Invention patent application deemed withdrawn after publication