WO2016095261A1

WO2016095261A1 - Method for estimating and restraining distributed iq imbalance

Info

Publication number: WO2016095261A1
Application number: PCT/CN2014/095118
Authority: WO
Inventors: 赵辉; 张�诚; 李斗; 赵玉萍; 李红滨
Original assignee: 北京大学; 北京三梯通网络技术有限公司; 北京瀚诺科技有限公司
Priority date: 2014-12-18
Filing date: 2014-12-26
Publication date: 2016-06-23
Also published as: CN104486283A; CN104486283B

Abstract

Disclosed is a method for estimating and restraining a distributed IQ imbalance. The method comprises: demodulating a received OFDM system signal, and outputting a training signal S(k) and a data signal Y(k) in a frequency domain; for the training signal S(k), removing modulation information in the training signal S(k) to obtain an interference item J(k) caused by IQ imbalance; then, for each subcarrier in a frame in which the training signal S(k) is located, estimating IQ imbalance parameters A(k) and B(k) caused by a transmit end and a receive end; updating IQ imbalance parameters Ã(k) and B͂(k) of a system according to A(k) and B(k); performing channel equalization on a frequency domain signal Y(k) to acquire a signal X^(k); and correcting the IQ imbalance for X^(k). The present invention has a wide application range, can estimate and restrain IQ imbalance related to or unrelated to a frequency in a radio frequency circuit at a transmit end or a receive end, can effectively prevent influence of frequency deviation, and has low calculation complexity.

Description

Distributed IQ imbalance estimation and suppression method

Technical field

The present invention belongs to the field of digital communications, and in particular, to an IQ (in-phase orthogonal branch) imbalance estimation and suppression method for an OFDM (Orthogonal Frequency Division Multiplexing) communication system.

Background technique

As the bandwidth of digital communication systems continues to increase, zero-IF RF transceiver solutions are increasingly used. However, IQ (in-phase orthogonal branch) imbalance is an important issue in the zero-IF scheme, which seriously affects the performance of the communication system. In the up-conversion or end-down conversion process of the originating end, there are two main factors that cause IQ imbalance: First, the two crystal oscillators are not completely orthogonal during the up-conversion process, which causes IQ imbalance unrelated to the frequency point. Second, the real (I-way) and imaginary (Q-way) signals pass through the amplifiers, filters and other analog devices are not completely consistent, this factor causes the frequency imbalance-related IQ imbalance.

In zero-IF systems, especially those that also employ OFDM and high-order constellation modulation, the IQ imbalance problem has been studied more and more. IQ imbalance can cause mutual interference between OFDM mirror subcarriers, affecting the correctness of the constellation demodulation process. However, the existing IQ imbalance estimation and suppression methods have many limitations. For example, most algorithms only study the single-ended IQ imbalance problem at the origin or the end; most algorithms only study the IQ independent of the frequency. Balance problem; most algorithms do not consider the influence of carrier frequency deviation on IQ imbalance estimation and suppression; most algorithms have large computational complexity, which is not conducive to hardware implementation.

Summary of the invention

The purpose of the present invention is to avoid the limitations of the above prior methods, and to propose a distributed IQ imbalance estimation and suppression method in an OFDM system. The invention has wide application range, and can estimate and suppress the IQ imbalance related to or not related to the frequency point in the RF circuit at the originating or terminating end, and at the same time, can effectively resist the influence of the frequency offset and has low computational complexity.

The technical solution of the present invention includes the following steps:

a) performing OFDM demodulation on the received training signal in the first frame to obtain a frequency domain signal S(k), where k is an OFDM subcarrier number,

N is the total number of subcarriers; removing the modulation information in S(k) results in the interference term J(k) caused by IQ imbalance, J(k)=S(k)-S ₀ (k)H(k), where H(k) is the channel response on the kth subcarrier of the lth frame, and S ₀ (k) is the modulation information on the kth subcarrier of the lth frame;

b) for each subcarrier, estimate the IQ imbalance parameters caused by the originating and terminating ends, namely A(k) and B(k),

Where m=round(2f _Δ ), that is, 2 times the relative frequency offset f _△ is rounded off;

c) According to the estimated IQ imbalance parameters A(k) and B(k) in frame l, the system IQ imbalance parameter

with

Update,

Where δ is a constant greater than 0 and less than 1;

d) performing OFDM demodulation on the received data signal in the first frame to obtain a frequency domain signal Y(k), and performing channel equalization to obtain a signal

Correct

Correction of IQ imbalance is obtained

Sending to the subsequent module to complete the demodulation of the signal, wherein the symbol '*' represents a conjugate operation on the variable;

e) The receiver continuously receives more frames, l+1th frame, l+2th frame, ..., and repeats the above 4 steps.

In the steps a) and d), the received first frame includes two parts of the training signal and the data signal. However, the present invention is not limited to such a frame structure. For example, in some communication systems, the training signal and the data signal are two independent frames, and the present invention is equally applicable to such a frame structure.

In the step a), the modulation information S ₀ (k) may be obtained by decision feedback, or may be information known at both ends of the transmission and reception; the channel response H(k) may be obtained by various channel estimation algorithms.

In the step b), the relative frequency offset f _Δ can be obtained by various frequency offset estimation algorithms.

The updating method in the step c) is an exponentially weighted moving average, wherein δ is a constant, and the larger the δ is, the faster the estimation speed of the IQ imbalance is; the smaller the δ is, the higher the estimation accuracy of the IQ imbalance is. In this step, other update methods, such as arithmetic averaging, can be used.

Where l is a variable, representing that one frame containing training symbols has been received.

The above method gives the IQ imbalance estimation and suppression method for both the receiving end and the originating end. The above method can be applied to the IQ imbalance estimation and suppression at the receiving end, and also to the IQ imbalance estimation and suppression at the origin. For example, in step b), only A(k) or B(k) is estimated. After iterative update, perform step d).

Subtract the corresponding item when making corrections

or

The beneficial effects of the invention are:

1. The present invention makes full use of the time-varying slow characteristic of IQ imbalance. Through multi-frame distributed estimation, higher IQ imbalance estimation accuracy can be obtained, thereby improving the suppression effect of IQ imbalance.

2. The invention estimates the IQ imbalance parameters of the originating end and the receiving end in the frequency domain, and can be adapted to the IQ imbalance scene related to the frequency point or uncorrelated with the frequency point, and the present invention considers the frequency offset to the IQ imbalance. The impact of estimation and suppression is more in line with the actual application scenario.

3. The implementation complexity of the present invention is low, and only a small number of multiplication and division operations are implemented by hardware.

DRAWINGS

Figure 1 shows two examples of frame structures suitable for use in the present invention;

(a) a frame comprising a training signal and a data signal in two parts;

(b) a scene in which the training signal and the data signal are located in different frames;

Figure 2 is a flow chart of the algorithm of the present invention;

Figure 3 is a performance simulation diagram of the present invention.

detailed description

The distributed IQ imbalance estimation and suppression method of the present invention will be described in detail below with reference to the accompanying drawings, but does not constitute a limitation of the present invention.

FIG. 1 shows two examples of frame structure scenarios to which the present invention is applicable. In scenario (a), one frame includes two parts of a training signal and a data signal; in scene (b), there are two types of frames, respectively containing training signals. Or data signal.

2 is an algorithm flow of the present invention. Referring to FIG. 2, the implementation flow of the IQ imbalance estimation and suppression method of the present invention is as follows:

The received signal is first subjected to OFDM demodulation, that is, an FFT operation is performed; according to different frame structures, the training signal S(k) and the data signal Y(k) in the frequency domain are outputted for estimation of IQ imbalance and data information, respectively. Transfer.

The training signal S(k) is used for the estimation of the IQ imbalance parameter. First, the modulation information is removed from S(k) to obtain the interference term J(k) caused by the IQ imbalance; then the origin and the interference are calculated according to the interference term J(k). The IQ imbalance parameter of the receiving end completes the IQ imbalance estimation of the frame; finally, the frame estimates the IQ imbalance estimation result to store the final IQ imbalance parameter A(k) in the system and

Update.

The data signal Y(k) needs to suppress the IQ imbalance in the demodulation process, firstly equalize Y(k); then calculate the interference size of IQ imbalance from the equalized data and the IQ imbalance parameter stored in the system. Finally, the interference size is subtracted from the equalized data, and the IQ imbalance is suppressed and sent to the subsequent demodulation module.

FIG. 3 is a performance simulation diagram of the present invention. The simulation uses 4096QAM, and it can be seen that when the system has IQ imbalance, the system will not work (the square mark is a solid line in the figure), and the bit error rate is high, and the IQ proposed by the present invention is adopted. After the unbalanced estimation and suppression algorithm, the performance is significantly improved (the circle marked with a solid line in the figure), compared with the case where the IQ imbalance does not exist (the unmarked solid line in the figure), the present invention has only 1 dB of performance after correction. loss.

The specific embodiments of the present invention have been described in detail with reference to the preferred embodiments of the present invention. All modifications, equivalent substitutions, improvements, etc., made within the spirit and scope of the invention are intended to be included within the scope of the invention.

Claims

A distributed IQ imbalance estimation method, characterized in that the received OFDM signal is demodulated, and the training signal S(k) in the frequency domain is output, and k is an OFDM subcarrier number.
N is the total number of subcarriers; then the modulation information in the training signal S(k) is removed, and the interference term J(k) caused by the IQ imbalance is obtained; then, for each subcarrier in the frame of the training signal S(k), Estimating the IQ imbalance parameter A(k) caused by the origin and the IQ imbalance parameter B(k) caused by the termination, where
H(k) is the channel response on the kth subcarrier in the frame where the training signal S(k) is located, and S 0 (k) is the modulation information on the kth subcarrier in the frame where the training signal S(k) is located, m=round (2f △ ), that is, the rounding of the relative frequency offset f △ of 2 times; then the IQ imbalance parameter of the system according to the IQ imbalance parameter A(k)
Update, according to the IQ imbalance parameter B(k), the IQ imbalance parameter of the system
Update.
The method of claim 1 wherein said interference term J(k) = S(k) - S 0 (k) H(k).
Method according to claim 1 or 2, characterized in that
IQ imbalance parameters for OFDM systems
Update; according to the formula
IQ imbalance parameters for OFDM systems
An update is made; where δ is a constant greater than 0 and less than one.
A distributed IQ imbalance estimation and suppression method, characterized in that a received OFDM system signal is demodulated, and a frequency domain training signal S(k) and a data signal Y(k) are output; k is an OFDM subcarrier Numbering,
N is the total number of subcarriers of the frame in which the training signal S(k) is located;

For the training signal S(k) in each frame, the modulation information in the training signal S(k) is removed, and the interference term J(k) caused by the IQ imbalance is obtained; then the frame of the training signal S(k) is located For each subcarrier, estimate the IQ imbalance parameter A(k) caused by the origin and the IQ imbalance parameter B(k) caused by the termination, where

H(k) is the channel response on the kth subcarrier in the frame where the training signal S(k) is located, and S 0 (k) is the modulation information on the kth subcarrier in the frame where the training signal S(k) is located, m=round (2f △ ), that is, the rounding of the relative frequency offset f △ of 2 times; then the IQ imbalance parameter of the system according to the IQ imbalance parameter A(k)
Update, according to the IQ imbalance parameter B(k), the IQ imbalance parameter of the system
Update;

Channel equalization for the frequency domain signal Y(k) in each frame to obtain a signal
Then right
Correction of IQ imbalance:
The method according to claim 4, wherein said training signal S(k) and data signal Y(k) are located in the same frame, or said training signal S(k) and data signal Y(k) are respectively Located in different frames.
The method according to claim 4 or 5, characterized in that said interference term J(k) = S(k) - S 0 (k) H(k).
Method according to claim 4 or 5, characterized in that
IQ imbalance parameters for OFDM systems
Update; according to the formula
IQ imbalance parameters for OFDM systems
An update is made; where δ is a constant greater than 0 and less than one.