CN111726306B - Full duplex system phase noise suppression method based on two-stage adaptive filtering - Google Patents

Full duplex system phase noise suppression method based on two-stage adaptive filtering Download PDF

Info

Publication number
CN111726306B
CN111726306B CN202010391010.8A CN202010391010A CN111726306B CN 111726306 B CN111726306 B CN 111726306B CN 202010391010 A CN202010391010 A CN 202010391010A CN 111726306 B CN111726306 B CN 111726306B
Authority
CN
China
Prior art keywords
signal
adaptive filter
time
phase noise
frequency domain
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.)
Active
Application number
CN202010391010.8A
Other languages
Chinese (zh)
Other versions
CN111726306A (en
Inventor
管鹏鑫
禹宏康
赵玉萍
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.)
Peking University
Original Assignee
Peking University
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 Peking University filed Critical Peking University
Priority to CN202010391010.8A priority Critical patent/CN111726306B/en
Publication of CN111726306A publication Critical patent/CN111726306A/en
Application granted granted Critical
Publication of CN111726306B publication Critical patent/CN111726306B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H04L25/024—Channel estimation channel estimation algorithms
    • H04L25/0256—Channel estimation using minimum mean square error criteria
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03012—Arrangements for removing intersymbol interference operating in the time domain
    • H04L25/03019—Arrangements for removing intersymbol interference operating in the time domain adaptive, i.e. capable of adjustment during data reception
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006—Arrangements for removing intersymbol interference
    • H04L25/03159—Arrangements for removing intersymbol interference operating in the frequency domain

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Noise Elimination (AREA)

Abstract

The invention discloses a full duplex system phase noise suppression method based on two-stage adaptive filtering. The method comprises the following steps: 1) all-purposeThe first adaptive filter of the duplex node transmits a signal x (n) and an expected signal r (n) to complete time domain channel estimation, and an estimation value is obtained
Figure DDA0002485586650000011
Then calculating an error signal
Figure DDA0002485586650000012
And performing time domain self-interference elimination; then removing CP and converting the signal into frequency domain to obtain frequency domain signal E1(k) (ii) a 2) Full duplex node slave E1(k) Obtaining the frequency domain signal X (k) of the emission signal x (n), calculating
Figure DDA0002485586650000013
The second adaptive filter of the full-duplex node is then based on A (k), E1(k) Obtaining a phase noise frequency domain estimation value delta (k); then according to E1(k) Delta (k) to obtain an error signal E2(k) And phase noise suppression is completed. The self-interference elimination capability of the scheme has higher performance gain, and the phase noise is restrained.

Description

Full duplex system phase noise suppression method based on two-stage adaptive filtering
Technical Field
The invention belongs to the technical field of communication, and particularly relates to a full duplex system phase noise suppression method based on two-stage adaptive filtering.
Background
To alleviate the contradiction between the shortage of spectrum resources and the increasing demand for bandwidth, research on Full Duplex (FD) technology capable of improving spectrum efficiency is being conducted. Compared with the existing Frequency-Division Duplexing (FDD) and Time-Division Duplexing (TDD) technologies, the full-duplex technology can theoretically achieve double spectrum efficiency, and has the advantages of solving the hidden terminal problem, improving relay communication efficiency, enhancing communication safety and the like. Currently, FD technology has become one of the key technologies of the fifth Generation communication technology (5th Generation, 5G). However, since the transmission and reception operate at the same time and the same frequency, the local receiver receives a locally transmitted signal replica, referred to as a Self-Interference Signal (SI). The strong self-interference signal can drown out the far-end useful signal, thereby making the useful signal unable to be demodulated. For example, in a Wireless local area network (WiFi) system, the SI signal is 90dB higher than the desired signal.
With the development of full duplex technology, a great number of self-interference cancellation mechanisms mainly including propagation domain cancellation, analog domain cancellation and digital domain cancellation exist. The propagation domain elimination is to make the signal reach the local receiving end by using the path loss and then undergo large attenuation; the Analog domain elimination is to reconstruct an SI signal in a radio frequency domain and subtract the SI signal from a received signal, so as to avoid saturation of an Analog-to-Digital Converter (ADC) module of a receiving end and reduce quantization noise; digital domain cancellation reconstructs the SI signal based on the estimated self-interference channel and the known transmitted signal and subtracts from the received signal. Due to technical limitations, propagation domain and analog domain cancellation cannot reduce the SI signal to noise level, and therefore, it is usually necessary to combine the three cancellation schemes.
With the intensive research on the self-interference cancellation mechanism, the researchers found that in addition to the linear channel, non-linear factors in the circuit, such as phase noise, Power Amplifier (PA) non-linear effect, and IQ imbalance, etc., in the actual system all cause the cancellation capability of the system to be reduced, wherein the most serious influence on the system is the phase noise. For an Orthogonal Frequency Division Multiplexing (OFDM) system, Phase noise may generate Common Phase Error (CPE) and subcarrier Interference (ICI), and if not suppressed, the self-Interference cancellation capability of the system may be greatly reduced, thereby hindering the practical application of the full-duplex technology. One existing suppression scheme for phase noise is only applicable to a common oscillator and does not consider a far-end useful signal, and another scheme estimates a phase noise coefficient by using Minimum Mean Square Error (MMSE), and channel state information needs to be completely known, which is difficult in a practical system. While the traditional scheme reconstructs an SI signal to be eliminated based on an estimated channel and a transmitted signal, channel and phase noise in an actual system have time-varying characteristics, and a self-adaptive filter can well track channel changes thanks to a rapidly developed digital signal processing technology.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention aims to provide a full duplex system phase noise suppression method based on two-stage adaptive filtering.
The invention firstly establishes a receiving and transmitting signal model when phase noise exists in an Orthogonal Frequency Division Multiplexing (OFDM) system, and finds out that a Frequency domain self-interference signal is represented as linear convolution of the signal and the Frequency domain phase noise through theoretical analysis. Considering that the self-interference channel and the phase noise in an actual system both have time-varying characteristics, the invention provides a two-stage adaptive filter structure for realizing digital domain self-interference elimination in a phase noise scene. Finally, performance simulation is performed on the scenarios of the common oscillator and the separate oscillator in the OFDM system. The result shows that compared with the existing scheme, the self-interference elimination capability of the scheme has higher performance gain, and the influence of phase noise on the system is inhibited.
The technical scheme of the invention is as follows:
a full duplex system phase noise suppression method based on two-stage adaptive filtering comprises the following steps:
1) the first adaptive filter of the full-duplex node completes time domain channel estimation according to the transmitting signal x (n) and the expected signal r (n) at the time n to obtain an estimated value
Figure BDA0002485586630000021
Then calculating an error signal
Figure BDA0002485586630000022
Will error signal e1(n) sending the data to a decoder for judgment decoding to complete time domain self-interference elimination; then removing the cyclic prefix CP of the signal after the time domain self-interference elimination and converting the obtained signal into a frequency domain to obtain a frequency domain signal E1(k) (ii) a k is 0,1, … N-1, N is the sub-carrier number of OFDM system;
2) the full-duplex node receives the frequency domain signal E1(k) Obtaining the frequency domain signal X (k) of the emission signal x (n), calculating
Figure BDA0002485586630000023
The second adaptive filter of the full-duplex node is then based on A (k), E1(k) Obtaining a phase noise frequency domain estimation value delta (k); then according to E1(k) Delta (k) to obtain an error signal E2(k) Finishing the suppression of phase noise; wherein
Figure BDA0002485586630000024
Is composed of
Figure BDA0002485586630000025
In the frequency domain.
Further, the full-duplex node uses separate antennas, that is, the transmit chain and the receive chain use different antennas.
Further, an estimated value is obtained
Figure BDA0002485586630000026
The method comprises the following steps:
11) initializing tap coefficients w of a first adaptive filter1(0)＝0T；
12) Calculating the output signal y of the first adaptive filter at time n1(n)＝w1(n)Tx (n); wherein, w1(n) is the tap coefficient vector of the adaptive Filter at time n of Filter1, and x (n) is the input signal vector at time n of the first adaptive Filter;
13) calculating an error signal e at time n1(n)＝r(n)-y1(n); where r (n) is the expected signal at time n of the first adaptive filter, y1(n) is the output value of the first adaptive filter at time n;
14) updating tap coefficients
Figure BDA0002485586630000031
Wherein, mu1Is a set step size, and 0<μ<2，
Figure BDA0002485586630000032
Is the conjugate of x (n);
15) if the algorithm is not converged, returning to the step 12); if convergence, the updated tap coefficient is used as an estimated value
Figure BDA0002485586630000033
Further, the method for estimating the phase noise frequency domain value δ (k) comprises the following steps:
21) initializing tap coefficient w of second adaptive filter2(0)＝0T；
22) Calculating the output signal y of the second adaptive filter at time n2(n)＝w2(n)TA (n); wherein, w2(n) is the tap coefficient vector of the adaptive filter at time n of the second adaptive filter, and A (n) is the input signal vector at time n of the second adaptive filter;
23) calculating an error signal E at time n2(n)＝E1(n)-y2(n); wherein E1(n) is the expected signal at time n of the second adaptive filter;
24) updating tap coefficients
Figure BDA0002485586630000034
Wherein, mu2Is a set step size, and 0<μ<2，
Figure BDA0002485586630000035
Is the conjugate of A (n);
25) if the algorithm is not converged, returning to the step 22); if converging, the updated tap coefficient w2As an estimate of the phase noise frequency domain δ (k).
Further, according to formula E2(k)＝E1(k)-A(k)*w2Obtaining the k sub-carrier signal E after self-interference elimination2(k)。
The invention has the beneficial effects that:
1) establishing and analyzing a model and influence of phase noise in an OFDM full-duplex system;
2) the ICI generated by the phase noise is analyzed to obtain the conclusion that the ICI is expressed in the form of linear convolution of a signal and frequency domain phase noise, and a frequency domain suppression scheme based on an adaptive filter is provided;
3) by combining the requirements of channel estimation and self-interference elimination, a full-duplex self-interference elimination structure based on two-stage self-adaptive filtering is designed to realize dynamic tracking of channel and phase noise, so that the influence of the phase noise is inhibited, and the self-interference elimination capability of the system is improved;
4) through simulation verification, when the common oscillator is adopted, the self-interference elimination capability of the scheme is 2.5dB higher than that of the traditional scheme; when the split oscillator is adopted, the self-interference elimination capability of the scheme is 6dB higher than that of the traditional scheme.
Drawings
FIG. 1 is a diagram of self-interference cancellation for a full-duplex OFDM system;
FIG. 2 is a diagram of an adaptive filter self-interference cancellation scheme;
FIG. 3 is a two-stage self-interference cancellation architecture diagram;
FIG. 4 is a plot of cancellation capability versus Interference-to-Noise Ratio (INR);
FIG. 5 is a plot of cancellation capability versus phase noise 3dB bandwidth;
figure 6 is a plot of the cancellation capability versus the order of Filter 2.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following examples in the accompanying drawings. It should be understood that the specific examples described herein are intended to be illustrative only and are not intended to be limiting.
The invention researches a scene that two full-duplex nodes A and B based on OFDM system transmission, and only one node, namely the node A, is selected for model establishment and algorithm research because the hardware structures of the two nodes are symmetrical. The system framework of the node is shown in fig. 1. It should be noted that the node uses separate antennas, that is, the transmit chain and the receive chain use different antennas. The oscillators shown in fig. 1 are separate oscillators, and the case of a common oscillator will also be analyzed below. The digital cancellation may be performed in the time domain or in the frequency domain. Fig. 1 is a simple overview of the system operation flow.
The node a OFDM system has N subcarriers, x (k), k is 0,1, … N-1 is OFDM frequency domain data, and x (k) obtains a discrete time domain sequence x (N) after Inverse Fast Fourier Transform (IFFT) and Cyclic Prefix (CP) addition. Converting the discrete time domain sequence x (n) into a continuous signal x (t) by a Digital-to-Analog Converter (DAC) module, and further obtaining a radio frequency signal by up-conversion frequency of an oscillator
Figure BDA0002485586630000046
(note: here power amplifiers at the transmitting and receiving ends are omitted), i.e.
Figure BDA0002485586630000041
Then, the radio frequency signal of the A node passes through a self-interference channel hSI(t) to the receiving end of node A, where hSI(t) comprises a direct path and a far-end obstacle reflection path between the transceiving antennas, and meanwhile, because the two nodes work at the same time and the same frequency, the receiving signal of the A node receiving antenna also comprises a far-end useful OFDM time domain signal r transmitted by the B nodeu(t)。
Analog cancellation works primarily in the radio frequency domain, based on the principle of generating a replica of the SI signal and subtracting it from the received signal over a series of delay lines, each with different attenuation and loss, which can be equated to one channel. Analog cancellation can therefore be viewed as locally transmitting a signal through the corresponding channel. In conclusion, the invention can obtain the signal after analog elimination
Figure BDA0002485586630000042
Is composed of
Figure BDA0002485586630000043
Wherein, h (t) is a joint equivalent channel of the self-interference channel and the analog domain cancellation channel, and represents convolution operation.
Figure BDA0002485586630000044
And obtaining a baseband signal r (t) after down-conversion of the oscillator at the receiving end, and further obtaining a baseband digital signal r (n) through an ADC module.
Figure BDA0002485586630000045
Wherein, tdThe number of samples corresponding to the time required for propagation and self-interference cancellation in the analog domain, w (n) is white Gaussian noise, θT(n) and thetaR(n) is the discrete sampling point of the phase noise of the transmitting end and the receiving end, the phase noise of the invention adopts the common free oscillator model, and the modeling is the wiener process thetaT(n1)-θT(n2)～N(0,4πf3dB|n1-n2|Ts) That is, the phase difference value of two discrete samples obeys a mean value of 0 and a variance of 4 pi f3dB|n1-n2|TsWherein N represents a normal distribution, f3dB3dB bandwidth, T, for phase noisesIs the sampling time interval. If the system employs a Common Oscillator (Common Oscillator), then θT(n)＝θR(n-td) If a separation Oscillator (separator) is used, then θT(n) and thetaRAnd (n) are mutually independent random processes.
After removing the CP and FFT operation, a frequency domain representation of the received signal can be obtained.
Figure BDA0002485586630000051
Where H (k) is the frequency domain form of the time domain corresponding to h (n) of the self-interference channel, Ru(k) In the frequency domain of the useful signal, w (k) is in the frequency domain of gaussian noise. Deltak-lThe frequency-domain form of the phase noise being combined for the transmitting and receiving ends, i.e.
Figure BDA0002485586630000052
Discrete fourier transform of (d).
Figure BDA0002485586630000053
It can be known from the formula (2.4) that the existence of the phase noise generates a common phase error CPE and an inter-subcarrier interference ICI, which seriously hinder the self-interference cancellation capability of the system, and therefore the self-interference cancellation mechanism needs to suppress the influence of the phase noise.
Cancellation scheme
NLMS adaptive filter interference cancellation principle
Adaptive filters have found widespread use in the field of digital signal processing, and interference cancellation is one of its important applications. The adaptive filter can be characterized as a Finite tap Filter (FIR), and the tap coefficients can be dynamically adjusted according to an algorithm, as shown in fig. 2. Where s (n) is the input signal vector at time n of the adaptive filter, z (n) is the expected signal at time n of the adaptive filter, and w (n) [ w ]1,w2…wL]TIs the tap coefficient vector of the filter at time n, and L is the number of taps. y (n) ═ w (n)Ts (n) is the output signal of the filter at time n, whereTRepresenting the transpose of the vector. And e (n) z (n) -y (n) is an error signal. Finally, the system updates the filter tap coefficient w through a certain updating algorithm.
The output of the adaptive filter can be regarded as the input signal s passing through a linear channel w, and the principle of the adaptive filter is to remove the part of the expected signal that is highly linearly related to the input signal, then as can be seen from equation (2.4), if the phase noise is not considered and the receiving end signal r (n) is taken as the expected signal, the local transmitting signal x (n) is taken as the expected signal, when the adaptive algorithm converges, the filter tap w is the estimated amount of the self-interference channel h (n), and the filter output signal y (n) is the reconstructed SI signal, then the error signal is outputNumber being useful signal ru(n) mixing with noise. Finally, the error signal is only needed to be sent to a decoder for decision decoding. To this end, the adaptive filter performs the function of self-interference cancellation.
Common adaptive algorithms include Least Mean Square (LMS) and Recursive Least Squares (RLS) algorithms, and since the RLS algorithm needs to invert a matrix in updating tap coefficients, the operation complexity is high, while the LMS algorithm is simple and feasible, but since the step length of the LMS algorithm in updating the tap coefficients is a fixed value, the convergence speed is affected, so that the invention adopts a variable-step Normalized Least Mean Square (NLMS) algorithm. The algorithm updating steps are as follows:
1. tap coefficient w (0) of initialized adaptive filter is 0T；
2. Calculating the output signal y (n) w (n) of the adaptive filter at time nTs (n); w (n) is a tap coefficient vector of the adaptive filter at the moment n, and s (n) is an input signal vector of the adaptive filter at the moment n;
3. calculating an error signal e (n) at time n, z (n) -y (n); z (n) is the expected signal at time n of the adaptive filter;
4. updating tap coefficients
Figure BDA0002485586630000061
Wherein mu is a set step length and 0<μ<2，
Figure BDA0002485586630000062
Is the conjugate of s (n).
5. If the algorithm is not converged, the second step is returned until the convergence is reached. Wherein the convergence condition is that the tap coefficient w hardly changes any more with iteration.
2. System cancellation architecture
As can be seen from the formula (2.4), the influence of the phase noise appears in the frequency domain as a convolution of the phase noise frequency domain form δ and X (k) H (k), i.e. the phase noise frequency domain form
Figure BDA0002485586630000063
Thus, it can be seen that in the frequency domain, the signal x (k) h (k) passes through the linear channel δ (k). Combining the principle of adaptive filter interference cancellation, it can be known that since x (k) is known, if the channel h (k) is also known, the adaptive filter can be used to dynamically cancel the phase noise interference in the frequency domain.
In practice, the channel state information h (k) may not be completely known, but is estimated through a certain algorithm, and in consideration of the time-varying characteristics of the channel state and the phase noise, the present invention designs a two-stage adaptive self-interference cancellation architecture, as shown in fig. 3.
The first stage is to complete the time domain channel estimation and eliminate the CPE caused by phase noise. The input of the adaptive Filter1 is a local known transmission signal x (n), the desired signal is a reception signal r (n), and r (n) includes a self-interference signal, a far-end useful signal and noise as shown in equation (2.3). As can be seen from the foregoing adaptive Filter principle, when the algorithm converges, the tap coefficient of the adaptive Filter1 is the CPE equivalent time domain channel δ generated by the phase noise0Estimated value of h (n)
Figure BDA0002485586630000064
And error signal
Figure BDA0002485586630000065
I.e. time domain self-interference cancellation is completed. The CP is then removed and converted into a frequency domain-derived frequency domain signal E1(k) From the formula (2.4), E1(k) The expression is as follows.
Figure BDA0002485586630000066
And in the second stage, ICI elimination caused by phase noise is completed. The locally transmitted frequency domain signal x (k) is known and the channel estimate is obtained from the first stage, and thus can be obtained
Figure BDA0002485586630000067
Is composed of
Figure BDA0002485586630000068
The ICI of the phase noise is linear convolution of a (k) and the frequency domain form δ (k) of the phase noise, as shown in equation (2.4). Therefore, the input signal of the adaptive Filter2 is A (k), and the desired signal is E1(k) In which E1(k) Including ICI, the far-end desired signal, and gaussian white noise. When the algorithm converges, the tap coefficient of the Filter2 is the estimated value of delta (k), and the error signal E2(k) That is, the frequency domain signal after ICI is further removed based on the first stage CPE removal, and if a plurality of interference is completely eliminated, E2(k) The OFDM demodulation can be directly carried out for the far-end useful signals and the noise.
To this end, self-interference cancellation of the received signal is accomplished through two-stage adaptive filter processing, including linear cancellation (h (n)) and phase noise-induced non-linear cancellation (CPE and ICI).
Simulation result
In order to verify the cancellation scheme, the invention performs simulation analysis, which mainly includes different Interference-to-Noise ratios (INRs), 3dB bandwidths, adaptive filter orders, and the like. INR is defined as the ratio of self-interference signal power to noise power, i.e.
Figure BDA0002485586630000071
Wherein E isCPECPE Power generated for phase noise, EICIFor ICI power, EwIs the noise power. The invention is based on OFDM system simulation, wherein the number of subcarriers N is 2048, the bandwidth of the subcarriers is 15kHz, and the sampling period is Ts＝3.3×10-8And s, the data information is modulated by 16 QAM. The multi-path channel attenuation is 0dB, -5dB, -15dB, -20dB, and the time delay is 0, 3, 4 and 5 sampling points respectively.
The conventional approach is to estimate the channel, reconstruct the SI signal and cancel it from the received signal. In order to maintain fairness, if the present invention is not specifically described, the channel estimation value in the conventional scheme is the tap coefficient when the adaptive filter converges in the first stage of the scheme. In addition, since the MMSE scheme in the literature assumes that the channel is perfectly known, in order to ensure simulation fairness, the channel estimation value is assumed in the MMSE scheme simulation as a tap coefficient when the adaptive filter in the first stage of the scheme converges.
The evaluation index of the invention is the system interference elimination capability G.
Figure BDA0002485586630000072
Wherein E isIRepresenting the energy of the self-interference signal before cancellation, ErRepresenting the residual interference signal energy after self-interference cancellation, and N is the noise energy. And G represents the self-interference energy ratio before and after the system is eliminated, and the interference elimination capability of the system is embodied.
1. Relationship between Elimination Capacity and INR
This section mainly simulates the effect of INR on the cancellation capability. Set up f3dBThe self-interference cancellation capability of the present solution and the conventional solution under different INRs is simulated, and the simulation result is shown in fig. 4, where the SNR is 15dB, the order of Filter1 is 8, and the order of Filter2 is 50. As can be seen from fig. 4, the cancellation capability of the three schemes is enhanced as INR is increased, because when the system transmit power is constant, the lower INR is, the higher the system noise energy is, which results in larger errors of the channel estimation value and the phase noise coefficient estimation value, so that the reconstructed SI signal is different from the real SI signal by a larger amount, resulting in a reduction in the cancellation capability. In addition, the scheme and the MMSE scheme are superior to the traditional scheme under two scenes of sharing and separating oscillators.
Note that in the shared oscillator scenario, when INR is low, the performance of the scheme is slightly lower than that of the MMSE scheme, because the phase noise at the receiving end is a time delay copy of the phase noise at the transmitting end, the influence of the phase noise on the system is low at this time, and when INR is low, gaussian noise is large, so that the influence of the gaussian noise is dominant, while the MMSE scheme utilizes a priori statistics of white gaussian noise, and better considers the influence of the noise, so that the cancellation capability is high. When INR is increased and white Gaussian noise is reduced, the influence of phase noise is dominant, and the elimination capability of the scheme is superior to that of an MMSE scheme.
For the separating oscillator, the transmitted and received phase noise are two independent random processes, and the influence on the system is much higher than that of gaussian noise, so the scheme is always better than the MMSE scheme.
Compared with the traditional scheme, the scheme has the advantages that the elimination capacity is about 2.5dB under the shared oscillator, the elimination capacity is 6dB under the separated oscillator, and the performance is improved.
2. Cancellation capability versus 3dB bandwidth
As can be seen from the phase noise model described above, the time-varying characteristic of the phase noise increases as the 3dB bandwidth increases, so the 3dB bandwidth has a certain effect on the self-interference cancellation capability.
The INR is set to be 50dB, the SNR is set to be 15dB, the order of Filter1 is 6, the order of Filter2 is 50, and the self-interference cancellation capability of the scheme, the MMSE scheme and the conventional scheme under different 3dB bandwidths is simulated. The simulation results are shown in fig. 5. As can be seen from fig. 5, the cancellation capability of all three schemes decreases as the 3dB bandwidth increases. The scheme and the MMSE scheme are superior to the traditional scheme. In the common oscillator scenario, when the coherence bandwidth f3dBThe MMSE scheme is slightly higher than the present scheme, when smaller. This is because when the coherence bandwidth is small, the phase noise of the common oscillator has less influence on the system than the gaussian noise, and therefore the MMSE scheme performance is slightly higher. And with the increase of the coherent bandwidth, the influence of the phase noise on the system gradually takes a dominant position, and the performance of the scheme is gradually superior to that of the MMSE scheme. In the scenario of a split oscillator, the phase noise is far higher than white gaussian noise, so the scheme is always better than the MMSE scheme.
3. Cancellation capability versus filter order
This section mainly studies the phase noise suppression capability of order M of Filter 2. Taking a split oscillator as an example, INR is set to 50dB, f3dB15Hz and SNR 15dB, respectively simulating the cancellation capability in three scenes of 10, 30 and 50, and the simulation result is shown in fig. 6. According to simulation results, as the order of the filter increases, the self-interference elimination capability of the system gradually increasesAnd (5) gradually strengthening. This is because the more filter tap coefficients, the more accurate the estimation of the phase noise, and therefore, the cancellation capability is improved.
Conclusion
The invention firstly establishes a mathematical model of the transmitting and receiving signals under the phase noise scene in the full-duplex OFDM system and analyzes the influence of the mathematical model. Compared with the traditional scheme, the scheme is based on an NLMS self-adaptive algorithm, can track the change of a channel and phase noise in real time, and has better practicability. When INR is high, the cancellation capability of this scheme is 2.5dB and 6dB higher than the conventional scheme with shared and split oscillators, respectively. In addition, as the 3dB bandwidth increases, the system cancellation capability decreases, but the scheme is always higher than the conventional scheme. As the filter order increases, the cancellation capability of the system also increases gradually.

Claims (4)

1. A full duplex system phase noise suppression method based on two-stage adaptive filtering comprises the following steps:
1) the first adaptive filter of the full-duplex node completes time domain channel estimation according to the transmitting signal x (n) and the expected signal r (n) at the time n to obtain an estimated value
Figure FDA0002969571210000011
Then calculating an error signal
Figure FDA0002969571210000012
Will error signal e1(n) sending the data to a decoder for judgment decoding to complete time domain self-interference elimination; then removing the cyclic prefix CP of the signal after the time domain self-interference elimination and converting the obtained signal into a frequency domain to obtain a frequency domain signal E1(k) (ii) a k is 0,1, … N-1, N is the sub-carrier number of OFDM system; wherein an estimated value is obtained
Figure FDA0002969571210000013
The method comprises the following steps: 11) initializing tap coefficients w of a first adaptive filter1(0)＝0T(ii) a 12) Calculating the output signal y of the first adaptive filter at time n1(n)＝w1(n)Tx (n); wherein, w1(n) is the tap coefficient vector of the adaptive filter at time n of the first adaptive filter, and x (n) is the input signal vector at time n of the first adaptive filter; 13) calculating an error signal e at time n1(n)＝r(n)-y1(n); where r (n) is the desired signal at time n of the first adaptive filter, y1(n) is the output value of the first adaptive filter at time n; 14) updating tap coefficients
Figure FDA0002969571210000014
Wherein, mu1Is a set step size, and 0<μ<2，
Figure FDA0002969571210000015
Is the conjugate of x (n); 15) if the algorithm is not converged, returning to the step 12); if convergence, the updated tap coefficient is used as an estimated value
Figure FDA0002969571210000016
2) The full-duplex node receives the frequency domain signal E1(k) Obtaining the frequency domain signal X (k) of the transmitting signal x (n), calculating the input signal of the second adaptive filter
Figure FDA0002969571210000017
The second adaptive filter of the full-duplex node is then based on A (k), E1(k) Obtaining a phase noise frequency domain estimation value delta (k); then according to E1(k) Delta (k) to obtain an error signal E2(k) Finishing the suppression of phase noise; wherein
Figure FDA0002969571210000018
Is composed of
Figure FDA0002969571210000019
In the frequency domain.
2. The method of claim 1, wherein the full-duplex node employs separate antennas, i.e., different antennas are used for transmit and receive chains.
3. The method of claim 1, wherein the phase noise frequency domain estimate δ (k) is obtained by:
21) initializing tap coefficient w of second adaptive filter2(0)＝0T；
22) Calculating the output signal y of the second adaptive filter at time n2(n)＝w2(n)TA (n); wherein, w2(n) is the tap coefficient vector of the adaptive filter at time n of the second adaptive filter, and A (n) is the input signal vector at time n of the second adaptive filter;
23) calculating an error signal E at time n2(n)＝E1(n)-y2(n); wherein E1(n) is the expected signal at time n of the second adaptive filter;
24) updating tap coefficients
Figure FDA00029695712100000110
Wherein, mu2Is a set step size, and 0<μ<2，
Figure FDA00029695712100000111
Is the conjugate of A (n);
25) if the algorithm is not converged, returning to the step 22); if converging, the updated tap coefficient w2As an estimate of the phase noise frequency domain δ (k).
4. The method of claim 1, according to formula E2(k)＝E1(k)-A(k)*w2Obtaining the k sub-carrier signal E after self-interference elimination2(k)，w2Is the tap coefficient of the second adaptive filter.
CN202010391010.8A 2020-05-11 2020-05-11 Full duplex system phase noise suppression method based on two-stage adaptive filtering Active CN111726306B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010391010.8A CN111726306B (en) 2020-05-11 2020-05-11 Full duplex system phase noise suppression method based on two-stage adaptive filtering

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010391010.8A CN111726306B (en) 2020-05-11 2020-05-11 Full duplex system phase noise suppression method based on two-stage adaptive filtering

Publications (2)

Publication Number Publication Date
CN111726306A CN111726306A (en) 2020-09-29
CN111726306B true CN111726306B (en) 2021-06-29

Family

ID=72566099

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010391010.8A Active CN111726306B (en) 2020-05-11 2020-05-11 Full duplex system phase noise suppression method based on two-stage adaptive filtering

Country Status (1)

Country Link
CN (1) CN111726306B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113315531B (en) 2021-05-25 2022-04-08 之江实验室 Method for receiving full duplex signals at same time and same frequency
CN115021885B (en) * 2022-05-19 2024-03-12 中国电子科技集团公司第三十六研究所 In-band full duplex system and interference signal cancellation method thereof

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102739578A (en) * 2011-03-31 2012-10-17 广东明家科技股份有限公司 Echo interference suppression method, device and terminal

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7003099B1 (en) * 2002-11-15 2006-02-21 Fortmedia, Inc. Small array microphone for acoustic echo cancellation and noise suppression
CN101320996B (en) * 2008-05-27 2012-10-10 中山大学 Self-adapting noise elimination apparatus and method
CN104539341B (en) * 2014-12-26 2018-02-13 西安电子科技大学 The active echo self-interference suppressing method of wireless broadband multi-antenna MIMO full duplex systems
CN104468055B (en) * 2014-12-29 2017-12-01 西安电子科技大学 Broadband wireless full duplex MIMO communication system echo self-interference Adaptive Suppression method
CN106534008B (en) * 2016-11-25 2019-08-02 西安烽火电子科技有限责任公司 The power compensation MMSE equalization methods of wireless multi-path channels
CN107659527B (en) * 2017-09-19 2020-09-08 电子科技大学 Phase noise suppression system and method for simultaneous same-frequency full duplex cooperative communication
US10374844B1 (en) * 2018-03-08 2019-08-06 Nxp B.V. Signal-equalization with noise-whitening filter

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102739578A (en) * 2011-03-31 2012-10-17 广东明家科技股份有限公司 Echo interference suppression method, device and terminal

Also Published As

Publication number Publication date
CN111726306A (en) 2020-09-29

Similar Documents

Publication Publication Date Title
CN108111186B (en) A kind of digital self-interference removing method of zero intermediate frequency full-duplex transceiver
US7609789B2 (en) Phase noise compensation for MIMO WLAN systems
US6996199B2 (en) Approach for processing data received from a communications channel to reduce noise power and optimize impulse response length to reduce inter-symbol interference and inter-channel interference
US7324437B1 (en) Method for co-channel interference cancellation in a multicarrier communication system
Hamazumi et al. A study of a loop interference canceller for the relay stations in an SFN for digital terrestrial broadcasting
US8605837B2 (en) Adaptive frequency-domain reference noise canceller for multicarrier communications systems
CN111726306B (en) Full duplex system phase noise suppression method based on two-stage adaptive filtering
CN101262253B (en) Method for solving cross-interference issue in digital user line and cross-interference elimination device
CA2381314A1 (en) Approach for processing data received from a communications channel in finite precision arithmetic applications
CN111092833B (en) Full-duplex underwater acoustic communication expectation channel and self-interference channel joint estimation method
Zhu et al. Blind compensation of frequency-dependent I/Q imbalance in direct conversion OFDM receivers
Achra et al. Performance analysis of MIMO OFDM system for different modulation schemes under various fading channels
JP2001244855A (en) Dsl transmission system which removes far-end crosstalk
WO2016067675A1 (en) Phase noise compensation receiver
CN112383322A (en) Regularization-based full-duplex system combined self-interference elimination method and electronic device
CN113872898B (en) Digital domain self-interference suppression method and system based on bounded component analysis
KR100738350B1 (en) Apparatus and Method of Equalization for Phase Noise Compensation in the Orthogonal Frequency Division Multiplexing communication system
CN112262533A (en) Active interference cancellation device, signal isolation control device and method for actively cancelling interference
WO2018000914A1 (en) Single-carrier channel estimation method
Zekkari et al. Efficient adaptive filtering algorithm for IQ imbalance compensation Tx/Rx systems<? show [AQ ID= Q1]?>
Liu et al. Digital nonlinear self-interference cancellation based on LMS-Volterra algorithm
JP4358706B2 (en) Interference canceller and relay apparatus using the interference canceller
Fouladifard et al. On adaptive cancellation of IQ mismatch in OFDM receivers
Anttila et al. Efficient mitigation of frequency-selective I/Q imbalance in OFDM receivers
Suyama et al. Decision-directed phase noise compensation for millimeter-wave single carrier systems with iterative frequency-domain equalization

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
GR01 Patent grant
GR01 Patent grant