TWI794001B - Radio frequency self-interference elimination method for full-duplex wireless receiver - Google Patents

Radio frequency self-interference elimination method for full-duplex wireless receiver Download PDF

Info

Publication number
TWI794001B
TWI794001B TW111103830A TW111103830A TWI794001B TW I794001 B TWI794001 B TW I794001B TW 111103830 A TW111103830 A TW 111103830A TW 111103830 A TW111103830 A TW 111103830A TW I794001 B TWI794001 B TW I794001B
Authority
TW
Taiwan
Prior art keywords
channel
signal
filter response
channel signal
auxiliary
Prior art date
Application number
TW111103830A
Other languages
Chinese (zh)
Other versions
TW202332227A (en
Inventor
鄧俊宏
曾怡雰
Original Assignee
元智大學
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 元智大學 filed Critical 元智大學
Priority to TW111103830A priority Critical patent/TWI794001B/en
Application granted granted Critical
Publication of TWI794001B publication Critical patent/TWI794001B/en
Publication of TW202332227A publication Critical patent/TW202332227A/en

Links

Images

Abstract

本發明為一種全雙工無線接收機之射頻消除自我干擾方法，包括建立一射頻不完美聯合訊號模型。輸入主路訊號與輔路訊號至射頻不完美聯合訊號模型中，以合成產生不完美輸出訊號。根據射頻不完美聯合訊號模型及不完美輸出訊號，對主路訊號與該輔路訊號之通道濾波響應進行估測，以取得主路訊號及輔路訊號之通道濾波響應參數。根據主路訊號及輔路訊號之通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，並估算測出預補償參數，以補償不完美輸出訊號。本發明能估測出訊號中的預補償參數，以解決收發機訊號洩漏的問題，提升訊號傳遞效益。The invention is a radio frequency self-interference elimination method of a full-duplex wireless receiver, which includes establishing a radio frequency imperfect combined signal model. Input the main channel signal and the auxiliary channel signal to the RF imperfect joint signal model to synthesize an imperfect output signal. According to the RF imperfect joint signal model and the imperfect output signal, estimate the channel filter response of the main channel signal and the auxiliary channel signal, so as to obtain the channel filter response parameters of the main channel signal and the auxiliary channel signal. According to the channel filter response parameters of the main channel signal and the auxiliary channel signal, the channel filter response of the RF imperfect joint signal model is compensated, and the pre-compensation parameters are estimated to compensate for the imperfect output signal. The invention can estimate the pre-compensation parameters in the signal, so as to solve the problem of signal leakage of the transceiver and improve the efficiency of signal transmission.

Description

全雙工無線接收機之射頻消除自我干擾方法Radio frequency self-interference elimination method for full-duplex wireless receiver

本發明係有關一種訊號傳輸技術，特別是指一種全雙工無線接收機之射頻消除自我干擾方法。 The invention relates to a signal transmission technology, in particular to a radio frequency self-interference elimination method of a full-duplex wireless receiver.

訊號傳輸時，相同的頻帶上若同時進行發送和接收，會導致嚴重的自我干擾，因此為解決上述問題，現今的無線通訊都是以半雙工的形式傳送。但無線全雙工傳送可提升更高的頻譜使用效率，因此如何消除全雙工無線接收機的自我干擾，對於全雙工無線接收機來說是至關重要的問題。 During signal transmission, if simultaneous transmission and reception are performed on the same frequency band, it will cause serious self-interference. Therefore, in order to solve the above problems, today's wireless communications are transmitted in the form of half-duplex. However, wireless full-duplex transmission can improve higher spectrum utilization efficiency, so how to eliminate the self-interference of the full-duplex wireless receiver is a crucial issue for the full-duplex wireless receiver.

無線傳收機在同時發射與接收訊號時，自身發射訊號環回進入接收端，其中微波發射和反射是自我干擾引發的原因之一。近年來為了克服此問題，提出方法有天線分離、天線消除、類比消除和數位消除等方法。但上述這些方法，礙於天線分離與天線消除，易受到天線與天線間隔的距離，及可用物理設備的空間限制。 When the wireless transceiver transmits and receives signals at the same time, the self-transmitting signal loops back into the receiving end, and microwave transmission and reflection are one of the reasons for self-interference. In order to overcome this problem in recent years, methods such as antenna separation, antenna cancellation, analog cancellation and digital cancellation have been proposed. However, these methods mentioned above are easily limited by the distance between antennas and the space of available physical devices due to antenna separation and antenna elimination.

有鑑於此，本發明遂針對上述習知技術之缺失，提出一種全雙工無線接收機之射頻消除自我干擾方法，以有效克服上述之該等問題。 In view of this, the present invention proposes a radio frequency self-interference elimination method for a full-duplex wireless receiver to effectively overcome the above-mentioned problems.

本發明之主要目的在提供一種全雙工無線接收機之射頻消除自我干擾方法，其能估測出訊號中的通道濾波響應以及預補償參數，以解決收發機訊號洩漏的問題，提升收發機間的隔離度性能，提升訊號傳遞效益。 The main purpose of the present invention is to provide a radio frequency self-interference elimination method for a full-duplex wireless receiver, which can estimate the channel filter response and pre-compensation parameters in the signal, so as to solve the problem of transceiver signal leakage and improve the communication between transceivers. Excellent isolation performance to improve signal transmission efficiency.

本發明之另一目的在提供一種全雙工無線接收機之射頻消除自我干擾方法，其能估測出訊號中的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數的問題，以對訊號進行補償，提升訊號傳遞效益。 Another object of the present invention is to provide a radio frequency self-interference elimination method for a full-duplex wireless receiver, which can estimate the amplitude (In-phase, I) imbalance parameter and phase (Quadrature, Q) imbalance in the signal The problem of parameters is used to compensate the signal and improve the efficiency of signal transmission.

為達上述之目的，本發明係提供一種全雙工無線接收機之射頻消除自我干擾方法，包括下列步驟，建立一射頻不完美聯合訊號模型。輸入主路訊號與輔路訊號至射頻不完美聯合訊號模型中，以將主路訊號與輔路訊號合成，產生不完美輸出訊號。根據射頻不完美聯合訊號模型及不完美輸出訊號，利用最小平方法或最小均方法對主路訊號與輔路訊號之通道濾波響應進行估測，以取得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數。根據主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，以估算測出預補償參數。利用預補償參數補償不完美輸出訊號。 In order to achieve the above purpose, the present invention provides a radio frequency self-interference elimination method for a full-duplex wireless receiver, which includes the following steps to establish a radio frequency imperfect combined signal model. Input the main channel signal and the auxiliary channel signal to the RF imperfect joint signal model to synthesize the main channel signal and the auxiliary channel signal to generate an imperfect output signal. According to the RF imperfect joint signal model and the imperfect output signal, use the least square method or the least average method to estimate the channel filter response of the main channel signal and the auxiliary channel signal, so as to obtain the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal. Channel filter response parameters. According to the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, the channel filter response of the radio frequency imperfect joint signal model is compensated to estimate and measure the pre-compensation parameters. Compensate for imperfect output signals using precompensation parameters.

在本實施例中，射頻不完美聯合訊號模型表示為：

Figure 111103830-A0305-02-0003-276
其中x M (n)為主路訊號，x A (n)為輔路訊號，c M (n)為主路訊號之通道濾波響應參數，c A (n)為輔路訊號之通道濾波響應參數，r(n)為不完美輸出訊號。 In this embodiment, the RF imperfect joint signal model is expressed as:
Figure 111103830-A0305-02-0003-276
 Among them, x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, c M ( n ) is the channel filter response parameter of the main channel signal, c A ( n ) is the channel filter response parameter of the auxiliary channel signal, r ( n ) is the imperfect output signal.

在本實施例中，利用最小平方法或最小均方法對主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，係使用最小均方法(Least Mean Square，LMS)對主路訊號與輔路訊號之通道濾波響應進行估測，包括下列步驟，首先將不完美輸出訊號帶入最小均方法式，表示為： J LMS =avg{e[n]e *[n]} 其中

Figure 111103830-A0305-02-0004-1
Figure 111103830-A0305-02-0004-275
為主路訊號之通道濾波響應參數及
Figure 111103830-A0305-02-0004-271
為輔路訊號之通道濾波響應參數；將
Figure 111103830-A0305-02-0004-272
及
Figure 111103830-A0305-02-0004-274
帶入遞迴關係式，表示為：
Figure 111103830-A0305-02-0004-3
 對遞迴關係式進行微分，以估測出主路訊號之通道濾波響應參數，及輔路訊號之通道濾波響應參數，遞迴關係式進行微分表示為：
Figure 111103830-A0305-02-0004-5
 In this embodiment, in the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal using the least square method or the least mean method, the least mean method (Least Mean Square, LMS) is used to estimate the channel filter response of the main channel signal and the auxiliary channel signal. Estimating the channel filter response of the auxiliary channel signal includes the following steps. Firstly, the imperfect output signal is brought into the least mean method, expressed as: J LMS = avg { e [ n ] e * [ n ]} where
Figure 111103830-A0305-02-0004-1
Figure 111103830-A0305-02-0004-275
The channel filter response parameters of the main channel signal and
Figure 111103830-A0305-02-0004-271
is the channel filter response parameter of the auxiliary channel signal;
Figure 111103830-A0305-02-0004-272
and
Figure 111103830-A0305-02-0004-274
Bringing in the recursive relation, expressed as:
Figure 111103830-A0305-02-0004-3
 Differentiate the recursive relationship to estimate the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal. The differential expression of the recursive relationship is expressed as:
Figure 111103830-A0305-02-0004-5

在本實施例中，根據主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，並估算測出預補償參數之步驟更包括，加入預補償代數至射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型。帶入主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數至理想射頻聯合訊號模型中，以估測出預補償參數。 In this embodiment, according to the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, the step of compensating the channel filter response of the radio frequency imperfect joint signal model, and estimating and measuring the precompensation parameters further includes: Compensate algebraically to the RF imperfect joint signal model to create an ideal RF joint signal model. Bring the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal into the ideal radio frequency joint signal model to estimate the pre-compensation parameters.

在本實施例中，理想射頻聯合訊號模型表示為：

Figure 111103830-A0305-02-0004-280
其中x M (n)為主路訊號，x A (n)為輔路訊號，c M (n)為主路訊號之通道濾波響應參數，c A (n)為輔路訊號之通道濾波響應參數，R(n)為完美輸出訊號，q A (n)為預補償代數，
Figure 111103830-A0305-02-0005-265
為迴旋運算處理。 In this embodiment, the ideal RF joint signal model is expressed as:
Figure 111103830-A0305-02-0004-280
 Among them, x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, c M ( n ) is the channel filter response parameter of the main channel signal, c A ( n ) is the channel filter response parameter of the auxiliary channel signal, R ( n ) is the perfect output signal, q A ( n ) is the precompensation algebra,
Figure 111103830-A0305-02-0005-265
It is processed by convolution operation.

在本實施例中，帶入主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數至理想射頻聯合訊號模型中，以估測出預補償參數之步驟更包括：將

Figure 111103830-A0305-02-0005-266
轉換為矩陣形，表示為：c M =-C A q A q A 為q A (n)的向量表示，C A 為該c A (n)的常對角矩陣(Toeplitz)表示；使用偽逆矩陣(Pseudoinverse)轉換q A ，以估測出預補償參數，偽逆矩陣的q A 表示為：q A =-(C A H C A )-1 C A H c M 。 In this embodiment, the step of bringing the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal into the ideal radio frequency joint signal model to estimate the precompensation parameters further includes:
Figure 111103830-A0305-02-0005-266
Converted to a matrix form, expressed as: c M =- C A q A q A is the vector representation of q A ( n ), C A is the constant diagonal matrix (Toeplitz) representation of the c A ( n ); using pseudo-inverse The matrix (Pseudoinverse) transforms q A to estimate the pre-compensation parameters, and the q A of the pseudo-inverse matrix is expressed as: q A =-( C A H C A ) -1 C A H c M .

在本實施例中，在根據射頻不完美聯合訊號模型及不完美輸出訊號，對主路訊號與輔路訊號之通道濾波響應進行估測，以取得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數之步驟中，更包括對主路訊號與輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測。 In this embodiment, according to the RF imperfect joint signal model and the imperfect output signal, the channel filter responses of the main channel signal and the auxiliary channel signal are estimated to obtain the channel filter response parameters of the main channel signal and the channel of the auxiliary channel signal In the step of filtering response parameters, the amplitude (In-phase, I) unbalance parameter and phase (Quadrature, Q) unbalance parameter of the main channel signal and the auxiliary channel signal, and the amplitude channel of the main channel signal and the auxiliary channel signal are also included. The amplitude channel filter response parameters of the main channel signal and the phase channel filter response parameters of the phase channels of the main channel signal and the auxiliary channel signal are estimated.

在本實施例中，射頻不完美聯合訊號模型表示為：

Figure 111103830-A0305-02-0005-264
其中
Figure 111103830-A0305-02-0005-6
Figure 111103830-A0305-02-0006-7
 x M (n)為主路訊號，x A (n)為輔路訊號，r(n)為不完美輸出訊號，
Figure 111103830-A0305-02-0006-252
為主路訊號之振幅通道濾波響應參數，α M 為主路訊號的振幅不平衡參數，θ M 為主路訊號的相位不平衡參數，
Figure 111103830-A0305-02-0006-253
為主路訊號之相位通道濾波響應參數，
Figure 111103830-A0305-02-0006-254
為輔路訊號之振幅通道濾波響應參數，α A 為輔路訊號的相位不平衡參數，θ A 為輔路訊號的該相位不平衡參數，
Figure 111103830-A0305-02-0006-255
為輔路訊號之相位通道濾波響應參數。 In this embodiment, the RF imperfect joint signal model is expressed as:
Figure 111103830-A0305-02-0005-264
 in
Figure 111103830-A0305-02-0005-6
Figure 111103830-A0305-02-0006-7
 x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, r ( n ) is the imperfect output signal,
Figure 111103830-A0305-02-0006-252
The amplitude channel filter response parameter of the main channel signal, α M is the amplitude imbalance parameter of the main channel signal, θ M is the phase imbalance parameter of the main channel signal,
Figure 111103830-A0305-02-0006-253
The phase channel filter response parameters of the master signal,
Figure 111103830-A0305-02-0006-254
is the amplitude channel filter response parameter of the auxiliary channel signal, α A is the phase imbalance parameter of the auxiliary channel signal, θ A is the phase imbalance parameter of the auxiliary channel signal,
Figure 111103830-A0305-02-0006-255
It is the phase channel filter response parameter of auxiliary channel signal.

在本實施例中，對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用最小均方法(Least Mean Square，LMS)進行估測，包括下列步驟，將不完美輸出訊號帶入最小均方法式，表示為：J LMS =avg{e[n]e *[n]}其中

Figure 111103830-A0305-02-0006-8
h M,±(n)為主路訊號之通道濾波響應參數，h A,±(n)為輔路訊號之通道濾波響應參數。將
Figure 111103830-A0305-02-0006-261
、
Figure 111103830-A0305-02-0006-263
、
Figure 111103830-A0305-02-0006-262
、
Figure 111103830-A0305-02-0006-259
帶入遞迴關係式，表示為：
Figure 111103830-A0305-02-0007-11
 對遞迴關係式進行微分，以估測出對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數，表示為：
Figure 111103830-A0305-02-0007-281
 In this embodiment, in the step of estimating the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal, the least average Method (Least Mean Square, LMS) to estimate, including the following steps, bringing the imperfect output signal into the least mean square method, expressed as: J LMS = avg { e [ n ] e * [ n ]} where
Figure 111103830-A0305-02-0006-8
 h M , ± ( n ) is the channel filter response parameter of the main channel signal, h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal. Will
Figure 111103830-A0305-02-0006-261
,
Figure 111103830-A0305-02-0006-263
,
Figure 111103830-A0305-02-0006-262
,
Figure 111103830-A0305-02-0006-259
Bringing in the recursive relation, expressed as:
Figure 111103830-A0305-02-0007-11
 Differentiate the recursive relation to estimate the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal, expressed as:
Figure 111103830-A0305-02-0007-281

在本實施例中，根據通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，並估算測出預補償參數之步驟更包括對主路訊號與輔路訊號的振幅通道的振幅預補償參數，以及相位通道的相位預補償參數進行估測，其步驟更包括，加入振幅預補償代數及相位預補償代數至射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型。帶入振幅通道濾波響應參數、相位通道濾波響應參數、振幅不平衡參數及相位不平衡參數至理想射頻聯合訊號模型中，以估測出振幅預補償參數及相位預補償參數。 In this embodiment, according to the channel filter response parameters, the channel filter response of the RF imperfect joint signal model is compensated, and the step of estimating and measuring the pre-compensation parameters further includes the amplitude pre-compensation parameters of the amplitude channels of the main channel signal and the auxiliary channel signal , and the phase precompensation parameters of the phase channel are estimated, and the steps further include adding amplitude precompensation algebra and phase precompensation algebra to the radio frequency imperfect joint signal model to establish an ideal radio frequency joint signal model. Bring the amplitude channel filter response parameters, phase channel filter response parameters, amplitude imbalance parameters and phase imbalance parameters into the ideal radio frequency joint signal model to estimate the amplitude precompensation parameters and phase precompensation parameters.

在本實施例中，理想射頻聯合訊號模型表示為：

Figure 111103830-A0305-02-0007-282
其中
Figure 111103830-A0305-02-0008-283
 其中h M,±(n)為主路訊號之通道濾波響應參數，h A,±(n)為輔路訊號之通道濾波響應參數，w 1(n)為振幅預補償代數，w 2(n)相位預補償代數，R(n)為完美輸出訊號。 In this embodiment, the ideal RF joint signal model is expressed as:
Figure 111103830-A0305-02-0007-282
 in
Figure 111103830-A0305-02-0008-283
 Among them, h M , ± ( n ) is the channel filter response parameter of the main channel signal, h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal, w 1 ( n ) is the amplitude precompensation algebra, w 2 ( n ) Phase precompensation algebra, R ( n ) is the perfect output signal.

在本實施例中，帶入振幅通道濾波響應參數、相位通道濾波響應參數、振幅不平衡參數及相位不平衡參數至理想射頻聯合訊號模型中，以估測出振幅預補償參數及相位預補償參數之步驟更包括，將理想射頻聯合訊號模型轉換為矩陣，其表示為：

Figure 111103830-A0305-02-0008-284
其中H A,+與H A,-為h A,+(n)與h A,-(n)之托普利茲(Toeplitz)矩陣表示，w 1為振幅預補償參數，w 2為相位預補償參數；及採用最小平方法，並使用偽逆矩陣(Pseudoinverse)，求得w 1及w 2表示為：
Figure 111103830-A0305-02-0008-285
 其中H為H A,+與H A,-的合成矩陣，H +=(H H H)-1 H H 。 In this embodiment, the amplitude channel filter response parameters, phase channel filter response parameters, amplitude imbalance parameters and phase imbalance parameters are brought into the ideal RF joint signal model to estimate the amplitude precompensation parameters and phase precompensation parameters The step of further includes converting the ideal RF joint signal model into a matrix, which is expressed as:
Figure 111103830-A0305-02-0008-284
 Among them, H A ,+ and H A ,- are represented by the Toeplitz matrix of h A ,+ ( n ) and h A ,- ( n ), w 1 is the amplitude pre-compensation parameter, w 2 is the phase pre-compensation parameters; and adopt the least square method, and use the pseudo-inverse matrix (Pseudoinverse), to obtain w 1 and w 2 expressed as:
Figure 111103830-A0305-02-0008-285
 Where H is the composite matrix of H A ,+ and H A ,- , H + =( H H H ) -1 H H .

底下藉由具體實施例詳加說明，當更容易瞭解本發明之目的、技術內容、特點及其所達成之功效。 In the following detailed description by means of specific embodiments, it will be easier to understand the purpose, technical content, characteristics and effects of the present invention.

1:收發機 1: Transceiver

10:主路訊號發射器 10: Main road signal transmitter

12:輔路訊號發射器 12: Auxiliary road signal transmitter

14:參數估測器 14: Parameter Estimator

16:線性濾波器 16: Linear filter

18:切換器 18:Switcher

20:主路濾波器 20: Main road filter

22:輔路濾波器 22: Auxiliary channel filter

24:訊號接收器 24: Signal receiver

S10~S18:步驟 S10~S18: Steps

第一圖係為本發明方法應用之收發機系統示意圖。 The first figure is a schematic diagram of a transceiver system where the method of the present invention is applied.

第二圖係為本發明之方法流程圖。 The second figure is a flow chart of the method of the present invention.

第三圖係為本發明之訊號預補償前後比較頻譜圖。 The third figure is a comparison spectrum diagram before and after the signal pre-compensation of the present invention.

第四圖係為本發明另一實施例之訊號預補償前後比較頻譜圖。 The fourth figure is a comparison spectrum diagram before and after signal pre-compensation according to another embodiment of the present invention.

本實施例能估測出訊號中的通道濾波響應及預補償參數，以解決收發機訊號洩漏的問題，提升收發機間的隔離度性能，提升訊號傳遞效益。 This embodiment can estimate the channel filter response and pre-compensation parameters in the signal, so as to solve the problem of transceiver signal leakage, improve the isolation performance between transceivers, and improve the efficiency of signal transmission.

說明本實施例之方法如何達到上述之功效，首先請參照第一圖，以說明本實施例之方法所應用之收發機的系統架構圖，如圖所示，收發機1包括一主路訊號發射器10、一輔路訊號發射器12、一參數估測器14、一線性濾波器16、一切換器18、一主路濾波器20、一輔路濾波器22及一訊號接收器24。 Describe how the method of this embodiment achieves the above-mentioned effects. First, please refer to the first figure to illustrate the system architecture diagram of the transceiver used in the method of this embodiment. As shown in the figure, the transceiver 1 includes a main channel signal transmitter device 10 , an auxiliary signal transmitter 12 , a parameter estimator 14 , a linear filter 16 , a switcher 18 , a main filter 20 , an auxiliary filter 22 and a signal receiver 24 .

在本實施例中，主路訊號發射器10及輔路訊號發射器12輸出的訊號為相位偏移調變(Quadrature Phase-Shift Keying，QPSK)訊號。主路濾波器20設有消除主路訊號之通道濾波響應的通道濾波響應參數c M (n)。輔路濾波器22設有消除輔路訊號之通道濾波響應的通道濾波響應參數c A (n)。參數估測器14為處理器，用以估測主路濾波器20及輔路濾波器22的通道濾波響應參數c M (n)、c A (n)及線性濾波器16內的預補償代數q A (n)。 In this embodiment, the signals output by the main channel signal transmitter 10 and the auxiliary channel signal transmitter 12 are Quadrature Phase-Shift Keying (QPSK) signals. The main channel filter 20 is provided with a channel filter response parameter c M ( n ) for eliminating the channel filter response of the main channel signal. The auxiliary channel filter 22 is provided with a channel filter response parameter c A ( n ) for eliminating the channel filter response of the auxiliary channel signal. The parameter estimator 14 is a processor for estimating the channel filter response parameters c M ( n ) and c A ( n ) of the main filter 20 and the auxiliary filter 22 and the precompensation algebra q in the linear filter 16 A ( n ).

主路訊號發射器10發送已知訓練主路訊號x M (n)，主路訊號x M (n)帶有通道濾波響應參數c M (n)，透過主路濾波器20補償通道濾波響應參數c M (n)後發射出補償後的主路訊號x SI (n)。輔路訊號發射器12發出已知訓練碼輔路訊號x A (n)，輔路訊號x A (n)也帶有通道濾波響應參數c A (n)，透過輔路濾波器22補償通道濾波響應參數c A (n)後發射出補償後的輔路訊號

Figure 111103830-A0305-02-0009-248
。 The main channel signal transmitter 10 sends a known training main channel signal x M ( n ), the main channel signal x M ( n ) has a channel filter response parameter c M ( n ), and the channel filter response parameter is compensated through the main channel filter 20 After c M ( n ), the compensated main channel signal x SI ( n ) is transmitted. The auxiliary channel signal transmitter 12 sends out a known training code auxiliary channel signal x A ( n ), and the auxiliary channel signal x A ( n ) also has a channel filter response parameter c A ( n ), and the channel filter response parameter c A is compensated by the auxiliary channel filter 22 ( n ) and then send out the compensated auxiliary channel signal
Figure 111103830-A0305-02-0009-248
.

訊號接收器24接收洩漏的補償後主路訊號x SI (n)及補償後輔路訊號

Figure 111103830-A0305-02-0010-243
後混合。此時若補償後輔路訊號
Figure 111103830-A0305-02-0010-244
能剛好屬於補償後主路訊號x SI (n)的反向信號，則可完美的消除洩漏的訊號，但若補償後輔路訊號
Figure 111103830-A0305-02-0010-246
無法完美的消除洩漏的補償後主路訊號，就會產生不完美輸出訊號。因此，若此參數估測器14能估算出準確的預補償參數，將有助於補償後主路訊號完全消除洩漏回訊號接收器24的主路訊號補償後主路訊號x SI (n)，提升收發機的訊號收發性能。 The signal receiver 24 receives the leaked compensated main channel signal x SI ( n ) and the compensated auxiliary channel signal
Figure 111103830-A0305-02-0010-243
Post-mix. At this time, if the auxiliary road signal is compensated
Figure 111103830-A0305-02-0010-244
Can just belong to the reverse signal of the main channel signal x SI ( n ) after compensation, the leaked signal can be perfectly eliminated, but if the auxiliary channel signal after compensation
Figure 111103830-A0305-02-0010-246
The compensated main signal that cannot perfectly eliminate the leakage will produce an imperfect output signal. Therefore, if the parameter estimator 14 can estimate accurate pre-compensation parameters, it will help the compensated main channel signal to completely eliminate the main channel signal leaking back to the signal receiver 24. After compensation, the main channel signal x SI ( n ), Improve the signal sending and receiving performance of the transceiver.

收發機1應用在估測線性濾波器16的預補償代數q A (n)時，首先操作主路訊號發射器10發送主路訊號x M (n)，經過主路濾波器20濾除主路訊號的通道濾波響應後發射出。同時切換器18切換使輔路訊號發射器12發出輔路訊號x A (n)不經由線性濾波器16，直接經過輔路濾波器22濾除輔路訊號的通道濾波響應後發射出。此時部分主路訊號與輔路訊號洩漏至訊號接收器，令補償後主路訊號x SI (n)與補償後輔路訊號

Figure 111103830-A0305-02-0010-247
混合進入到參數估測器14，以估測出預補償代數q A (n)後，提供給線性濾波器16補償訊號。 When the transceiver 1 is used to estimate the pre-compensation algebra q A ( n ) of the linear filter 16, firstly, the main channel signal transmitter 10 is operated to send the main channel signal x M (n), and the main channel signal x M ( n ) is filtered by the main channel filter 20 The signal's channel filtered response is emitted. At the same time, the switch 18 is switched so that the auxiliary channel signal x A ( n ) sent by the auxiliary channel signal transmitter 12 does not pass through the linear filter 16 , but is directly transmitted through the auxiliary channel filter 22 to filter out the channel filter response of the auxiliary channel signal. At this time, part of the main channel signal and auxiliary channel signal leaks to the signal receiver, so that the compensated main channel signal x SI ( n ) and the compensated auxiliary channel signal
Figure 111103830-A0305-02-0010-247
The mixture enters the parameter estimator 14 to estimate the pre-compensation algebra q A ( n ), and then provides the compensation signal to the linear filter 16 .

當要對傳輸出去的訊號補償時，切換器18再進行切換，使輔路訊號進入到線性濾波器16，以補償輔路訊號，當輔路訊號輸出時，即能有效補償補償後主路訊號x SI (n)，以發射完美訊號。 When it is necessary to compensate the transmitted signal, the switcher 18 switches again so that the auxiliary channel signal enters the linear filter 16 to compensate the auxiliary channel signal. When the auxiliary channel signal is output, it can effectively compensate the compensated main channel signal x SI ( n ), to transmit a perfect signal.

接著說明本實施例全雙工無線接收機之射頻消除自我干擾方法，透過本方法能估測出主路濾波器20的主路訊號之通道濾波響應參數c M (n)，輔路濾波器22的輔路訊號之通道濾波響應參數c A (n)，以及線性濾波器16內的預補償代數q A (n)。全雙工無線接收機之射頻消除自我干擾方法詳述如下。 Next, the radio frequency self-interference elimination method of the full-duplex wireless receiver of the present embodiment is described. Through this method, the channel filter response parameter c M ( n ) of the main channel signal of the main channel filter 20, and the channel filter response parameter c M (n) of the auxiliary channel filter 22 can be estimated. The channel filter response parameter c A ( n ) of the auxiliary channel signal, and the pre-compensation algebra q A ( n ) in the linear filter 16 . The radio frequency self-interference elimination method of the full-duplex wireless receiver is detailed as follows.

首先進入步驟S10，建立一射頻不完美聯合訊號模型。接著進入步驟S12，輸入主路訊號與輔路訊號至參數估測器14的射頻不完美聯合訊號模型中，以將主路訊號與輔路訊號合成，產生不完美輸出訊號。其中射頻不完美聯合訊號模型表示為式(1)：

Figure 111103830-A0305-02-0011-241
其中x M (n)為主路訊號，x A (n)為輔路訊號，c M (n)為主路訊號之通道濾波響應參數，c A (n)為輔路訊號之通道濾波響應參數，r(n)為不完美輸出訊號，
Figure 111103830-A0305-02-0011-242
為迴旋運算處理。 First enter step S10, and establish a radio frequency imperfect joint signal model. Then enter step S12, input the main channel signal and the auxiliary channel signal into the radio frequency imperfect joint signal model of the parameter estimator 14, so as to synthesize the main channel signal and the auxiliary channel signal to generate an imperfect output signal. The radio frequency imperfect joint signal model is expressed as formula (1):
Figure 111103830-A0305-02-0011-241
 Among them, x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, c M ( n ) is the channel filter response parameter of the main channel signal, c A ( n ) is the channel filter response parameter of the auxiliary channel signal, r ( n ) is the imperfect output signal,
Figure 111103830-A0305-02-0011-242
It is processed by convolution operation.

接著進入步驟S14，參數估測器14先估測主路濾波器20的主路訊號之通道濾波響應參數c M (n)，輔路濾波器22的輔路訊號之通道濾波響應參數c A (n)。該步驟係根據射頻不完美聯合訊號模型及不完美輸出訊號，利用最小平方法或最小均方法對主路訊號與輔路訊號之通道濾波響應進行估測，以取得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數。在本實施例中，估測主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數可使用最小平方法(Least Square，LS)、最小均方法(Least Mean Square，LMS)、遞迴最小平方法(Recursive Least Squares，RLS)、歸一化最小均方法(Normalized Least Mean Squares，NLMS)或比例歸一化最小均方法(Proportionate Normalized Least Mean Square，PNLMS)等演算法進行估測。 Then enter step S14, the parameter estimator 14 first estimates the channel filter response parameter c M ( n ) of the main channel signal of the main channel filter 20, and the channel filter response parameter c A ( n ) of the auxiliary channel signal of the auxiliary channel filter 22 . This step is based on the RF imperfect joint signal model and the imperfect output signal, using the least square method or the least mean method to estimate the channel filter response of the main channel signal and the auxiliary channel signal, so as to obtain the channel filter response parameters of the main channel signal and The channel filter response parameter of auxiliary channel signal. In this embodiment, the least square method (Least Square, LS), the least mean method (Least Mean Square, LMS), the recursive most Recursive Least Squares (RLS), Normalized Least Mean Squares (NLMS) or Proportionate Normalized Least Mean Squares (PNLMS) and other algorithms are used for estimation.

使用最小平方法(Least Square，LS)估測主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數之步驟，詳述如下。以最小平方法(LS)來說，透過測量誤差平均值為0，使其計算出匹配的函數曲線並求得最小殘差平方和的總和。 The steps of estimating the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal using the least square method (Least Square, LS) are described in detail as follows. In terms of the least square method (LS), by measuring the average value of the error to be 0, it can calculate the matching function curve and obtain the sum of the minimum residual square sum.

將不完美輸出訊號r(n)轉換以矩陣r表示式(2)：

Figure 111103830-A0305-02-0012-17
其中X M 為x M (n)的常對角矩陣(Toeplitz)表示，X A 為x A (n)的常對角矩陣(Toeplitz)表示，c M 為c M (n)的向量表示，c A 為c A (n)的向量表示。並提出誤差值表示為式(3)：
Figure 111103830-A0305-02-0012-18
 Transform the imperfect output signal r ( n ) into matrix r expression (2):
Figure 111103830-A0305-02-0012-17
 Where X M is the constant diagonal matrix (Toeplitz) representation of x M ( n ), X A is the constant diagonal matrix (Toeplitz) representation of x A ( n ), c M is the vector representation of c M ( n ), c A is the vector representation of c A ( n ). And the error value is expressed as formula (3):
Figure 111103830-A0305-02-0012-18

接著利用最小平方法的代價函數式，估測主路訊號之通道濾波響應參數

Figure 111103830-A0305-02-0012-239
及輔路訊號之通道濾波響應參數
Figure 111103830-A0305-02-0012-278
。此法則代價函數表示為式(4)：
Figure 111103830-A0305-02-0012-20
 代價函數利用微分求濾波器響應值
Figure 111103830-A0305-02-0012-21
，如下式(5)：
Figure 111103830-A0305-02-0012-286
 Then use the cost function formula of the least square method to estimate the channel filter response parameters of the main channel signal
Figure 111103830-A0305-02-0012-239
And the channel filter response parameters of the auxiliary channel signal
Figure 111103830-A0305-02-0012-278
. The regular cost function is expressed as formula (4):
Figure 111103830-A0305-02-0012-20
 The cost function uses differentiation to find the filter response value
Figure 111103830-A0305-02-0012-21
, as formula (5):
Figure 111103830-A0305-02-0012-286

求得

Figure 111103830-A0305-02-0013-225
與
Figure 111103830-A0305-02-0013-226
為預估值
Figure 111103830-A0305-02-0013-277
與
Figure 111103830-A0305-02-0013-228
之向量，以帶入主路濾波器20及輔路濾波器22作為主路訊號之通道濾波響應參數c M (n)，輔路訊號之通道濾波響應參數c A (n)使用補償訊號。 obtain
Figure 111103830-A0305-02-0013-225
and
Figure 111103830-A0305-02-0013-226
is an estimated value
Figure 111103830-A0305-02-0013-277
and
Figure 111103830-A0305-02-0013-228
The vector is brought into the main channel filter 20 and the auxiliary channel filter 22 as the channel filter response parameter c M ( n ) of the main channel signal, and the channel filter response parameter c A ( n ) of the auxiliary channel signal uses the compensation signal.

說明使用最小均方法(Least Mean Square，LMS)估測主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數之步驟。詳細來說，相較於LS方法以固定的濾波器設計，在無線通訊中，對於估測誤差可隨環境與時間自動調整等化器參數，適應性等化器更能滿足此需求，在此說明最小均方法如何運用迭代的方式得出適應性濾波響應值。 Describe the steps of estimating the channel filter response parameters of the main channel signal and channel filter response parameters of the auxiliary channel signal using the Least Mean Square (LMS) method. In detail, compared with the LS method with a fixed filter design, in wireless communication, the parameters of the equalizer can be automatically adjusted with the environment and time for the estimation error, and the adaptive equalizer can better meet this demand. Here Explain how the least mean method uses an iterative method to obtain the adaptive filter response value.

首先將不完美輸出訊號帶入最小均方法式，表示為式(6)：J LMS =avg{e[n]e *[n]} (6)其中誤差值表示為式(7)：

Figure 111103830-A0305-02-0013-23
Figure 111103830-A0305-02-0013-229
為主路訊號之通道濾波響應參數及
Figure 111103830-A0305-02-0013-231
為輔路訊號之通道濾波響應參數。將
Figure 111103830-A0305-02-0013-232
及
Figure 111103830-A0305-02-0013-234
帶入遞迴關係式，表示為式(8)、式(9)：
Figure 111103830-A0305-02-0013-24
 First, the imperfect output signal is brought into the least mean method, which is expressed as formula (6): J LMS = avg { e [ n ] e * [ n ]} (6) where the error value is expressed as formula (7):
Figure 111103830-A0305-02-0013-23
Figure 111103830-A0305-02-0013-229
The channel filter response parameters of the main channel signal and
Figure 111103830-A0305-02-0013-231
It is the channel filter response parameter of auxiliary channel signal. Will
Figure 111103830-A0305-02-0013-232
and
Figure 111103830-A0305-02-0013-234
Bring in the recursive relationship, expressed as formula (8), formula (9):
Figure 111103830-A0305-02-0013-24

Figure 111103830-A0305-02-0013-25
對遞迴關係式(8)、式(9)進行微分，表示如下式(10)及式(11)：針對
Figure 111103830-A0305-02-0013-238
Figure 111103830-A0305-02-0014-27
 針對
Figure 111103830-A0305-02-0014-220
Figure 111103830-A0305-02-0014-28
 最後估測出主路訊號之通道濾波響應參數，及輔路訊號之通道濾波響應參數，遞迴關係式進行微分表示為式(12)：
Figure 111103830-A0305-02-0014-29
Figure 111103830-A0305-02-0013-25
 Differentiate the recursive relational formula (8) and formula (9), and express the following formula (10) and formula (11): for
Figure 111103830-A0305-02-0013-238
Figure 111103830-A0305-02-0014-27
 against
Figure 111103830-A0305-02-0014-220
Figure 111103830-A0305-02-0014-28
 Finally, the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal are estimated, and the recursive relationship is differentiated and expressed as formula (12):
Figure 111103830-A0305-02-0014-29

使用遞迴最小平方法(recursive least squares，RLS)估測主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數之步驟，係將不完美輸出訊號帶入遞迴最小平方法式，以求得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數。詳細來說，RLS方法以其出色的性能與更快的收斂性聞名，遞迴最小平方法也是一種適應性濾波器算法，利用已知迭代橫向濾波器係數計算下一個時刻的迭代橫向濾波器係數。 The step of estimating the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal using the recursive least squares method (recursive least squares, RLS) is to bring the imperfect output signal into the recursive least squares method to obtain The channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal are obtained. In detail, the RLS method is known for its excellent performance and faster convergence. The recursive least squares method is also an adaptive filter algorithm that uses known iterative transversal filter coefficients to calculate the iterative transversal filter coefficients at the next moment. .

遞迴最小平方法式之代價函數式表示為式(13)：

Figure 111103830-A0305-02-0014-123
上式(13)中，λ為介於0與1之間的遺忘因子，δ為正規化參數，根據訊雜比設置，當訊雜比大δ則小，訊雜比小則反之，
Figure 111103830-A0305-02-0015-212
與
Figure 111103830-A0305-02-0015-213
為預估的主路通道與輔路通道的通道濾波響應參數。其中誤差值e[n]表示為式(14)：
Figure 111103830-A0305-02-0015-121
 其中預估的主路訊號與輔路訊號的通道濾波響應參數
Figure 111103830-A0305-02-0015-214
與
Figure 111103830-A0305-02-0015-222
，用矩陣形式表示成
Figure 111103830-A0305-02-0015-32
，主路訊號x M (n)與輔路訊號x A (n)用矩陣形式表示成X=[x M (n)x A (n)]。 將代價函數微分後表示為式(15)：
Figure 111103830-A0305-02-0015-125
 將式(15)簡化成式(16)：
Figure 111103830-A0305-02-0015-126
 式(16)中，k(n)為增益向量，當e[n]收斂時，可求出
Figure 111103830-A0305-02-0015-219
,
Figure 111103830-A0305-02-0015-218
整理後如式(17)：
Figure 111103830-A0305-02-0015-127
 The cost function of the recursive least squares method is expressed as formula (13):
Figure 111103830-A0305-02-0014-123
 In the above formula (13), λ is a forgetting factor between 0 and 1, and δ is a normalization parameter, which is set according to the signal-to-noise ratio. When the signal-to-noise ratio is large, δ is small, and vice versa.
Figure 111103830-A0305-02-0015-212
and
Figure 111103830-A0305-02-0015-213
is the estimated channel filter response parameter of the main channel and auxiliary channel. Where the error value e [ n ] is expressed as formula (14):
Figure 111103830-A0305-02-0015-121
 The channel filter response parameters of the estimated main channel signal and auxiliary channel signal
Figure 111103830-A0305-02-0015-214
and
Figure 111103830-A0305-02-0015-222
, expressed in matrix form as
Figure 111103830-A0305-02-0015-32
, the main channel signal x M ( n ) and the auxiliary channel signal x A ( n ) are expressed in matrix form as X=[ x M ( n ) x A ( n )]. After differentiating the cost function, it is expressed as formula (15):
Figure 111103830-A0305-02-0015-125
 Simplify formula (15) into formula (16):
Figure 111103830-A0305-02-0015-126
 In formula (16), k ( n ) is the gain vector, when e [ n ] converges, it can be obtained
Figure 111103830-A0305-02-0015-219
,
Figure 111103830-A0305-02-0015-218
After sorting, it is as formula (17):
Figure 111103830-A0305-02-0015-127

同LMS方法，皆須配合環境的狀況手動調整找出合適的變數λ與δ。 Similar to the LMS method, it is necessary to manually adjust to find the appropriate variables λ and δ according to the environmental conditions.

說明使用歸一化最小均方法(Normalized Least Mean Squares，NLMS)估測主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數之步驟，係將不完美輸出訊號帶入歸一化最小均方式，以求得該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數。詳細步驟如下。 Explain the steps of using the Normalized Least Mean Squares (NLMS) method to estimate the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, which is to bring the imperfect output signal into the normalized least mean squares method to obtain the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal. The detailed steps are as follows.

由LMS方法得知，估測方法須手動調整變數μ，人工調整的參數較不易精確，因此提出NLMS方法使輸入訊號的功率歸一化，以確保演算法能穩定並解決此問題。 According to the LMS method, the estimation method needs to manually adjust the variable μ , and the manually adjusted parameters are not easy to be accurate. Therefore, the NLMS method is proposed to normalize the power of the input signal to ensure the stability of the algorithm and solve this problem.

NLMS的誤差值表示為式(18)：

Figure 111103830-A0305-02-0016-36
The error value of NLMS is expressed as formula (18):
Figure 111103830-A0305-02-0016-36

預估的主路通道與輔路通道的通道濾波響應參數

Figure 111103830-A0305-02-0016-210
與
Figure 111103830-A0305-02-0016-211
，用矩陣形式表示成
Figure 111103830-A0305-02-0016-38
，x M (n)與x A (n)用矩陣形式表示成X=[x M (n)x A (n)]。 Estimated channel filter response parameters of the main channel and auxiliary channel
Figure 111103830-A0305-02-0016-210
and
Figure 111103830-A0305-02-0016-211
, expressed in matrix form as
Figure 111103830-A0305-02-0016-38
, x M ( n ) and x A ( n ) are expressed in matrix form as X=[ x M ( n ) x A ( n )].

接著歸一化最小均方式之代價函數式，如式(19)：

Figure 111103830-A0305-02-0016-40
微分上式(19)，成為式(20)：
Figure 111103830-A0305-02-0016-41
 拉格朗日乘數λ，可表示為式(21)：
Figure 111103830-A0305-02-0017-44
 簡化歸一化最小均方式，表示為式(22)：
Figure 111103830-A0305-02-0017-45
 即可估計通道濾波響應參數，表示為式(23)：
Figure 111103830-A0305-02-0017-46
 Then normalize the cost function formula of the least average method, such as formula (19):
Figure 111103830-A0305-02-0016-40
 Differentiate the above formula (19) to become formula (20):
Figure 111103830-A0305-02-0016-41
 The Lagrangian multiplier λ can be expressed as formula (21):
Figure 111103830-A0305-02-0017-44
 Simplified normalized least mean method, expressed as formula (22):
Figure 111103830-A0305-02-0017-45
 The channel filter response parameters can be estimated, expressed as formula (23):
Figure 111103830-A0305-02-0017-46

使用比例歸一化最小均方法(proportionate normalized least mean square，PNLMS)對主路訊號與輔路訊號之通道濾波響應進行估測。比例歸一化最小均方法與NLMS相比，在反射波路徑稀疏時，PNLMS具有非常快的初始收斂與追蹤，若脈衝響應較為分散時，PNLMS的收斂速度就不及NLMS，由上述得知PNLMS易受到環境所影響。 歸一化最小均方法的代價函數表示為式(24)：

Figure 111103830-A0305-02-0017-128
上述式(24)將
Figure 111103830-A0305-02-0017-207
、r(n)、X(n)分成I與Q路表示，其中預估的主路通道與輔路通道的通道濾波響應參數
Figure 111103830-A0305-02-0017-208
與
Figure 111103830-A0305-02-0017-209
，用矩陣形式表示成
Figure 111103830-A0305-02-0017-48
，x M (n)與x A (n)用矩陣形式表示成X=[x M (n)x A (n)]。上式(24)歸一化最小均方法的代價函數經過微分後表示為式(25)：
Figure 111103830-A0305-02-0018-50
 拉格朗日乘數λ 1和λ 2，表示為式(26)：
Figure 111103830-A0305-02-0018-51
 上式(26)簡化，表示為式(27)：
Figure 111103830-A0305-02-0018-52
 其中
Figure 111103830-A0305-02-0018-53
，整理後如式(28)：P(n)=I-G(n-1)X(n)[X T (n)G(n-1)X(n)]-1 X T (n) (28)估計的主路通道與輔路通道的通道濾波響應參數表示式(29)：
Figure 111103830-A0305-02-0018-55
 其中
Figure 111103830-A0305-02-0018-56
The proportional normalized least mean square (PNLMS) method is used to estimate the channel filter response of the main channel signal and the auxiliary channel signal. Compared with NLMS, proportional normalized least mean method, PNLMS has very fast initial convergence and tracking when the reflected wave path is sparse. affected by the environment. The cost function of the normalized least mean method is expressed as formula (24):
Figure 111103830-A0305-02-0017-128
 The above formula (24) will be
Figure 111103830-A0305-02-0017-207
, r ( n ), X( n ) are divided into I and Q channels, where the estimated channel filter response parameters of the main channel and auxiliary channel
Figure 111103830-A0305-02-0017-208
and
Figure 111103830-A0305-02-0017-209
, expressed in matrix form as
Figure 111103830-A0305-02-0017-48
, x M ( n ) and x A ( n ) are expressed in matrix form as X=[ x M ( n ) x A ( n )]. The cost function of the normalized least mean method in the above formula (24) is expressed as formula (25) after differentiation:
Figure 111103830-A0305-02-0018-50
 The Lagrange multipliers λ 1 and λ 2 are expressed as formula (26):
Figure 111103830-A0305-02-0018-51
 The above formula (26) is simplified and expressed as formula (27):
Figure 111103830-A0305-02-0018-52
 in
Figure 111103830-A0305-02-0018-53
, after sorting, it is as formula (28): P ( n )=I- G ( n -1) X ( n )[ X T ( n ) G ( n -1) X ( n )] -1 X T ( n ) (28) The channel filter response parameter expression (29) of the estimated main channel and auxiliary channel:
Figure 111103830-A0305-02-0018-55
 in
Figure 111103830-A0305-02-0018-56

在上述步驟S14通過最小平方法、最小均方法、遞迴最小平方法、歸一化最小均方法或比例歸一化最小均方法估測出主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數後，即可提供主路濾波器20及輔路濾波器22使用，濾除主路訊號與輔路訊號中的通道濾波響應。 In the above step S14, the channel filter response parameters of the main channel signal and the channel filter of the auxiliary channel signal are estimated by the least square method, the least mean method, the recursive least square method, the normalized least mean method or the proportional normalized least mean method After responding to the parameters, it can be provided to the main channel filter 20 and the auxiliary channel filter 22 to filter the channel filter responses in the main channel signal and the auxiliary channel signal.

進入步驟S16估測預補償參數。步驟S16係參數估測器14根據主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，以估算測出預補償參數。 Go to step S16 to estimate the pre-compensation parameters. Step S16 is that the parameter estimator 14 compensates the channel filter response of the RF imperfect joint signal model according to the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, so as to estimate and measure the pre-compensation parameters.

詳細來說，參數估測器14在估算預補償參數時，先加入預補償代數至射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型。其中理想射頻聯合訊號模型表示為式(30)：

Figure 111103830-A0305-02-0019-205
其中x M (n)為主路訊號，x A (n)為輔路訊號，c M (n)為主路訊號之通道濾波響應參數，c A (n)為輔路訊號之通道濾波響應參數，R(n)為完美輸出訊號，q A (n)為預補償代數。 In detail, when estimating the precompensation parameters, the parameter estimator 14 first adds precompensation algebra to the RF imperfect joint signal model to establish an ideal RF joint signal model. The ideal RF joint signal model is expressed as formula (30):
Figure 111103830-A0305-02-0019-205
 Among them, x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, c M ( n ) is the channel filter response parameter of the main channel signal, c A ( n ) is the channel filter response parameter of the auxiliary channel signal, R ( n ) is the perfect output signal, q A ( n ) is the precompensation algebra.

接著帶入上述步驟S14估測出的主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數至理想射頻聯合訊號模型中，以估測出預補償參數。其演算步驟係將

Figure 111103830-A0305-02-0019-206
轉換為矩陣形，表示為式(31)：c M =-C A q A (31)q A 為q A (n)的向量表示，C A 為c A (n)的托普利茲(Toeplitz)矩陣表示。接著使用偽逆矩陣(Pseudoinverse)轉換q A ，以估測出預補償參數q A (n)，偽逆矩陣的q A 表示為式(32)：q A =-(C A H C A )-1 C A H c M (32) Then, the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal estimated in the above step S14 are brought into the ideal RF joint signal model to estimate the pre-compensation parameters. Its calculation steps will be
Figure 111103830-A0305-02-0019-206
Converted to matrix form, expressed as formula (31): c M =- C A q A (31) q A is the vector representation of q A ( n ), C A is the Toeplitz of c A ( n ) matrix representation. Then use the pseudo-inverse matrix (Pseudoinverse) to convert q A to estimate the pre-compensation parameter q A ( n ). The pseudo-inverse matrix q A is expressed as formula (32): q A =-( C A H C A ) - 1 C A H c M (32)

最後進入步驟S18，利用預補償參數補償不完美輸出訊號。透過上述方法即可得到主路訊號之通道濾波響應參數c M (n)，輔路訊號之通道濾波響應參數c A (n)以及預補償參數q A ，故本實施例可將上述補償參數提供給能消除射頻洩漏之收發機1，以有效補償不完美輸出訊號。 Finally, enter step S18, using the pre-compensation parameters to compensate the imperfect output signal. The channel filter response parameter c M ( n ) of the main channel signal, the channel filter response parameter c A ( n ) of the auxiliary channel signal and the pre-compensation parameter q A can be obtained through the above method, so this embodiment can provide the above compensation parameters to Transceiver 1 capable of eliminating radio frequency leakage to effectively compensate for imperfect output signals.

在補償不完美輸出訊號時，主路訊號發射器10發射主路訊號x M (n)經過主路濾波器20，以濾除主路訊號的通道濾波響應。同時輔路訊號發射器12發射輔路訊號x A (n)，此時切換器18切換輔路訊號x A (n)進入線性濾波器16，以透過線性濾波器16的預補償參數q A 來補償輔路訊號，其再進入到輔路濾波器22，以補償輔路訊號的通道濾波響應。因此帶有預補償參數的輔路訊號就能完全消除部分洩漏的主路訊號，達到主動射頻洩漏消除之技術。 When compensating the imperfect output signal, the main channel signal transmitter 10 transmits the main channel signal x M ( n ) through the main channel filter 20 to filter out the channel filter response of the main channel signal. At the same time, the auxiliary channel signal transmitter 12 transmits the auxiliary channel signal x A ( n ), and at this time the switcher 18 switches the auxiliary channel signal x A ( n ) into the linear filter 16 to compensate the auxiliary channel signal through the pre-compensation parameter q A of the linear filter 16 , which then enters the auxiliary channel filter 22 to compensate the channel filter response of the auxiliary channel signal. Therefore, the auxiliary channel signal with pre-compensation parameters can completely eliminate part of the leaked main channel signal, achieving the technology of active radio frequency leakage elimination.

請參照第三圖，其為經本實施例所應用之方法所產生的實驗數據比較頻譜圖，藉由10兆赫(MHz)的單載波訊號可從頻譜圖上直接觀測本實施例之方法所估測預補償參數q A 是否有效抑制洩漏的主路訊號，且可透過計算執行主動射頻消除前後強度比值。由第四圖能明顯看出預補償前洩漏的主路訊號明顯突出，但預補償後已完全被削除。 Please refer to the third figure, which is a comparison spectrogram of experimental data generated by the method applied in this embodiment. The 10 megahertz (MHz) single carrier signal can be directly observed from the spectrogram and estimated by the method of this embodiment Whether the pre-compensation parameter q A effectively suppresses the leaked main channel signal, and can be calculated by calculating the intensity ratio before and after active radio frequency cancellation. From Figure 4, it can be clearly seen that the leaked main channel signal is prominent before pre-compensation, but it has been completely eliminated after pre-compensation.

然而本發明除了上述實施例所示，可針對通道濾波響應及不完美輸出訊號進行補償之外，更可針對寬頻射頻不完美因此進行補償，在本實施例中，寬頻射頻不完美因子所指的是，當基頻訊號經由射頻發送時，需要經過振幅/相位(In-phase/Quadrature，I/Q)調變器將訊號載送至高頻，其伴隨著射頻元件與震盪電路的誤差，因此產生振幅不平衡及相位不平衡，也就是IQ不平衡。因此，本實施例更針對射頻具有IQ不平衡情境下，與通道濾波響應共存，此響應將會併入IQ不平衡進行聯合估測，詳述如下。 However, in addition to the above embodiment, the present invention can compensate for channel filter response and imperfect output signal, and can also compensate for broadband radio frequency imperfections. In this embodiment, the broadband radio frequency imperfection factor refers to Yes, when the baseband signal is sent through the radio frequency, it needs to pass through the amplitude/phase (In-phase/Quadrature, I/Q) modulator to carry the signal to the high frequency, which is accompanied by the error of the radio frequency component and the oscillator circuit, so Amplitude imbalance and phase imbalance are generated, that is, IQ imbalance. Therefore, this embodiment is more aimed at the coexistence of channel filter response and IQ imbalance in the situation where the radio frequency has IQ imbalance, and the response will be incorporated into the IQ imbalance for joint estimation, as detailed below.

本實施例所應用的系統架構以及主要步驟流程與上述實施例相同，故說明本實施例時，仍以第一圖與第二圖配合說明。唯一不同在於主路濾波器20與輔路濾波器22增加了補償訊號的主路訊號與輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數， 及主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數。 The system architecture and main steps of this embodiment are the same as those of the above-mentioned embodiments, so when describing this embodiment, the first and second figures are still used for illustration. The only difference is that the main channel filter 20 and the auxiliary channel filter 22 increase the amplitude (In-phase, I) imbalance parameter and the phase (Quadrature, Q) imbalance parameter of the main channel signal and the auxiliary channel signal of the compensation signal, And the amplitude channel filter response parameter of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameter of the phase channel of the main channel signal and the auxiliary channel signal.

本實施例在估測參數時亦與上述實施例相同，在參數估測器14中進行參數估測。首先進步驟S10，在參數估測器14中建立一射頻不完美聯合訊號模型。輸入主路訊號與輔路訊號至射頻不完美聯合訊號模型中，以將主路訊號與輔路訊號合成，產生不完美輸出訊號。其中射頻不完美聯合訊號模型表示為式(33)：

Figure 111103830-A0305-02-0021-200
其中
Figure 111103830-A0305-02-0021-57
 x M (n)為主路訊號，x A (n)為輔路訊號，r(n)為不完美輸出訊號，
Figure 111103830-A0305-02-0021-201
為主路訊號之振幅通道濾波響應參數，α M 為主路訊號的振幅不平衡參數，θ M 為主路訊號的相位不平衡參數，
Figure 111103830-A0305-02-0021-202
為主路訊號之相位通道濾波響應參數，
Figure 111103830-A0305-02-0021-203
為輔路訊號之振幅通道濾波響應參數，α A 為輔路訊號的振幅不平衡參數，θ A 為輔路訊號的相位不平衡參數，
Figure 111103830-A0305-02-0021-204
為輔路訊號之相位通道濾波響應參數。 This embodiment is also the same as the above embodiment when estimating parameters, and parameter estimation is performed in the parameter estimator 14 . First, proceed to step S10 , establish a radio frequency imperfect joint signal model in the parameter estimator 14 . Input the main channel signal and the auxiliary channel signal to the RF imperfect joint signal model to synthesize the main channel signal and the auxiliary channel signal to generate an imperfect output signal. The radio frequency imperfect joint signal model is expressed as formula (33):
Figure 111103830-A0305-02-0021-200
 in
Figure 111103830-A0305-02-0021-57
 x M ( n ) is the main channel signal, x A ( n ) is the auxiliary channel signal, r ( n ) is the imperfect output signal,
Figure 111103830-A0305-02-0021-201
The amplitude channel filter response parameter of the main channel signal, α M is the amplitude imbalance parameter of the main channel signal, θ M is the phase imbalance parameter of the main channel signal,
Figure 111103830-A0305-02-0021-202
The phase channel filter response parameters of the master signal,
Figure 111103830-A0305-02-0021-203
is the amplitude channel filter response parameter of the auxiliary channel signal, α A is the amplitude imbalance parameter of the auxiliary channel signal, θ A is the phase imbalance parameter of the auxiliary channel signal,
Figure 111103830-A0305-02-0021-204
It is the phase channel filter response parameter of auxiliary channel signal.

進入步驟S14，參數估測器14根據射頻不完美聯合訊號模型及不完美輸出訊號，利用最小平方法或最小均方法對主路訊號與輔路訊號之通道濾波響應進行估測，以取得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數。在本實施例中，參數估測器14更對主路訊號與輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及主路訊號與輔路訊號的振幅通道的振幅通道濾波響應 參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測。 Entering step S14, the parameter estimator 14 uses the least square method or the least mean method to estimate the channel filter response of the main channel signal and the auxiliary channel signal according to the radio frequency imperfect joint signal model and the imperfect output signal, so as to obtain the main channel signal The channel filter response parameters of the channel and the channel filter response parameters of the auxiliary channel signal. In the present embodiment, the parameter estimator 14 further analyzes the amplitude (In-phase, I) imbalance parameter and the phase (Quadrature, Q) imbalance parameter of the main channel signal and the auxiliary channel signal, and the parameters of the main channel signal and the auxiliary channel signal Amplitude Channel Filter Response for Amplitude Channel parameters, and the phase channel filter response parameters of the phase channels of the main channel signal and the auxiliary channel signal are estimated.

本實施例中估測主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數可使用最小平方法(LS)、最小均方法(LMS)、遞迴最小平方法(RLS)、歸一化最小均方法(NLMS)或比例歸一化最小均方法(PNLMS)進行估測。 In this embodiment, the least square method (LS) and the least average method can be used to estimate the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal. (LMS), recursive least squares (RLS), normalized least mean (NLMS) or proportional normalized least mean (PNLMS).

首先說明使用最小平方法(Least Square，LS)對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測之方法。 Firstly, the method of estimating the amplitude channel filter response parameters of the amplitude channels of the main channel signal and the auxiliary channel signal and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal by using the least square method (Least Square, LS) .

針對未知的主路通道與輔路通道的通道濾波響應參數h M,±(n)與h A,±(n)，將接收訊號r(n)改為矩陣r表示，採用最小平方法，以Least Square(LS)方法來說，透過測量誤差平均值為0，使其計算出匹配的函數曲線並求得最小殘差平方和的總和。將不完美輸出訊號轉換以矩陣表示為式(34)：

Figure 111103830-A0305-02-0022-59
其中X表示將主路通道與輔路通道Toeplitz矩陣X M 、X * M 、X A 、X * A 合併成X=[X M X * M X A X * A ]，h M,±為h M,±(n)的向量表示，h A,±為h A,±(n)的向量表示。 定義誤差值e表示為式(35)：
Figure 111103830-A0305-02-0023-61
 欲求得預估的主路通道與輔路通道的通道濾波響應參數，利用小平方法的代價函數估測，表示為式(36)：
Figure 111103830-A0305-02-0023-63
 微分式(36)，求通道濾波響應參數，表示為式(37)：
Figure 111103830-A0305-02-0023-64
 最後求得預估值
Figure 111103830-A0305-02-0023-198
與
Figure 111103830-A0305-02-0023-199
之向量，以求得主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數。 For the unknown channel filter response parameters h M , ± ( n ) and h A , ± ( n ) of the main channel and the auxiliary channel, the received signal r ( n ) is changed to a matrix r , and the least square method is used to represent For the Square (LS) method, the average value of the measurement error is 0, so that it can calculate the matching function curve and obtain the sum of the minimum residual square sum. Transform the imperfect output signal into a matrix as formula (34):
Figure 111103830-A0305-02-0022-59
 Wherein X means to merge the Toeplitz matrix X M , X * M , X A , X * A of the main channel and the auxiliary channel into X = [ X M X * M X A X * A ], h M , ± is h M , The vector representation of ± ( n ), h A ,± is the vector representation of h A , ± ( n ). Define the error value e as formula (35):
Figure 111103830-A0305-02-0023-61
 In order to obtain the estimated channel filter response parameters of the main channel and the auxiliary channel, the cost function estimation using the Xiaoping method is expressed as formula (36):
Figure 111103830-A0305-02-0023-63
 Differential formula (36), to find channel filter response parameters, expressed as formula (37):
Figure 111103830-A0305-02-0023-64
 The final estimate
Figure 111103830-A0305-02-0023-198
and
Figure 111103830-A0305-02-0023-199
to obtain the amplitude channel filter response parameters of the amplitude channels of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal.

使用最小均方法(Least Mean Square，LMS)對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測之方法詳述如下。 Using the Least Mean Square (LMS) method to estimate the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal. described below.

相較於LS方法以固定的濾波器設計，在無線通訊中，對於估測誤差可隨環境與時間自動調整等化器參數，適應性等化器更能滿足此需求，在此介紹最小平方法如何運用迭代的方式得出適應性濾波響應值。 Compared with the LS method with a fixed filter design, in wireless communication, the parameters of the equalizer can be automatically adjusted according to the environment and time for the estimation error, and the adaptive equalizer can better meet this demand. The least square method is introduced here. How to use the iterative method to obtain the adaptive filter response value.

最小均方法之代價函數如式(38)： J LMS =avg{e[n]e *[n]} (38)其中誤差值e[n]表示為式(39)：

Figure 111103830-A0305-02-0024-66
h M,±(n)為主路訊號之通道濾波響應參數，h A,±(n)為輔路訊號之通道濾波響應參數。上式(37)中，將
Figure 111103830-A0305-02-0024-193
、
Figure 111103830-A0305-02-0024-194
、
Figure 111103830-A0305-02-0024-195
、
Figure 111103830-A0305-02-0024-197
帶入遞迴關係式如式(40)：
Figure 111103830-A0305-02-0024-67
 The cost function of the least mean method is as formula (38): J LMS = avg { e [ n ] e * [ n ]} (38) where the error value e [ n ] is expressed as formula (39):
Figure 111103830-A0305-02-0024-66
 h M , ± ( n ) is the channel filter response parameter of the main channel signal, h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal. In the above formula (37), the
Figure 111103830-A0305-02-0024-193
,
Figure 111103830-A0305-02-0024-194
,
Figure 111103830-A0305-02-0024-195
,
Figure 111103830-A0305-02-0024-197
Bring in the recursive relation as formula (40):
Figure 111103830-A0305-02-0024-67

對上式(38)各代價函數微分，以估測出該主路訊號之該通道濾波響應參數，及該輔路訊號之該通道濾波響應參數，並簡化成式(41)：

Figure 111103830-A0305-02-0025-68
Differentiate the cost functions of the above formula (38) to estimate the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, and simplify them into formula (41):
Figure 111103830-A0305-02-0025-68

使用遞迴最小平方法(recursive least squares，RLS)對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測，係將不完美輸出訊號帶入遞迴最小平方法式，以求得主路訊號之通道濾波響應參數及該輔路訊號之該通道濾波響應參數。詳細來說，RLS方法以其出色的性能與更快的收斂性聞名，遞迴最小平方法也是一種適應性濾波器算法，利用已知迭代橫向濾波器係數計算下一個時刻的迭代橫向濾波器係數。 遞迴最小平方法之價函數如式(42)：

Figure 111103830-A0305-02-0025-70
上式(42)中，λ為介於0與1之間的遺忘因子，δ為正規化參數，其中誤差值e[n]表示為式(43)：
Figure 111103830-A0305-02-0026-72
 接著將預估主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數，用矩陣表示成
Figure 111103830-A0305-02-0026-73
，x M (n)、
Figure 111103830-A0305-02-0026-190
、x A (n)、
Figure 111103830-A0305-02-0026-191
用矩陣表示成
Figure 111103830-A0305-02-0026-74
將代價函數微分後如下式(44)：
Figure 111103830-A0305-02-0026-76
 簡化成式(45)：
Figure 111103830-A0305-02-0026-129
 式(45)的k(n)為增益向量，e[n]收斂可求得
Figure 111103830-A0305-02-0026-192
，表示為式(46)：
Figure 111103830-A0305-02-0026-131
 Using the recursive least squares method (recursive least squares, RLS) to estimate the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal, The imperfect output signal is brought into the recursive least square method to obtain the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal. In detail, the RLS method is known for its excellent performance and faster convergence. The recursive least squares method is also an adaptive filter algorithm that uses known iterative transversal filter coefficients to calculate the iterative transversal filter coefficients at the next moment. . The price function of the recursive least square method is as formula (42):
Figure 111103830-A0305-02-0025-70
 In the above formula (42), λ is a forgetting factor between 0 and 1, and δ is a regularization parameter, where the error value e [ n ] is expressed as formula (43):
Figure 111103830-A0305-02-0026-72
 Then, the channel filter response parameters of the estimated main channel signal and the channel filter response parameters of the auxiliary channel signal are expressed in a matrix as
Figure 111103830-A0305-02-0026-73
, x M ( n ),
Figure 111103830-A0305-02-0026-190
, x A ( n ),
Figure 111103830-A0305-02-0026-191
Expressed in a matrix as
Figure 111103830-A0305-02-0026-74
The cost function is differentiated as follows (44):
Figure 111103830-A0305-02-0026-76
 Simplified into formula (45):
Figure 111103830-A0305-02-0026-129
 k ( n ) in formula (45) is the gain vector, and e [ n ] converges to obtain
Figure 111103830-A0305-02-0026-192
, expressed as formula (46):
Figure 111103830-A0305-02-0026-131

使用歸一化最小均方法(Normalized Least Mean Squares，MLMS)對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測。其係將不完美輸出訊號帶入歸一化最小均方式，以求得主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數。詳細來說，由LMS方 法得知，估測方法須手動調整變數μ，人工調整的參數較不易精確，因此提出NLMS方法使輸入訊號的功率歸一化，以確保演算法能穩定並解決此問題。 Use the normalized least mean method (Normalized Least Mean Squares, MLMS) to estimate the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal. Measurement. It brings the imperfect output signal into the normalized minimum average method to obtain the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal. In detail, from the LMS method, it is known that the estimation method must manually adjust the variable μ , and the manually adjusted parameters are not easy to be accurate. Therefore, the NLMS method is proposed to normalize the power of the input signal to ensure the stability of the algorithm and solve this problem. .

NLMS的誤差值表示為式(47)：

Figure 111103830-A0305-02-0027-132
其中預估的主路與輔路通道濾波響應參數，用矩陣表示成
Figure 111103830-A0305-02-0027-80
，x M (n)、
Figure 111103830-A0305-02-0027-188
、x A (n)、
Figure 111103830-A0305-02-0027-189
用矩陣表示成
Figure 111103830-A0305-02-0027-81
 代價函數如式(48)：
Figure 111103830-A0305-02-0027-133
 微分式(48)後為式(49)：
Figure 111103830-A0305-02-0027-135
 拉格朗日乘數λ，如式(50)：
Figure 111103830-A0305-02-0028-84
 簡化式(50)，表示為式(51)：
Figure 111103830-A0305-02-0028-85
 最後估計通道濾波響應參數如式(52)
Figure 111103830-A0305-02-0028-137
 The error value of NLMS is expressed as formula (47):
Figure 111103830-A0305-02-0027-132
 Among them, the estimated channel filter response parameters of the main road and the auxiliary road are expressed in a matrix as
Figure 111103830-A0305-02-0027-80
, x M ( n ),
Figure 111103830-A0305-02-0027-188
, x A ( n ),
Figure 111103830-A0305-02-0027-189
Expressed in a matrix as
Figure 111103830-A0305-02-0027-81
 The cost function is as formula (48):
Figure 111103830-A0305-02-0027-133
 Differential formula (48) is followed by formula (49):
Figure 111103830-A0305-02-0027-135
 Lagrangian multiplier λ , such as formula (50):
Figure 111103830-A0305-02-0028-84
 Simplified formula (50), expressed as formula (51):
Figure 111103830-A0305-02-0028-85
 The final estimated channel filter response parameters are as formula (52)
Figure 111103830-A0305-02-0028-137

使用比例歸一化最小均方法對主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測。與NLMS相比，在反射波路徑稀疏時，PNLMS具有非常快的初始收斂與追蹤，若脈衝響應較為分散時，PNLMS的收斂速度就不及NLMS，由上述得知PNLMS易受到環境所影響。 比例歸一化最小均方法之代價函數表示為式(53)：

Figure 111103830-A0305-02-0028-87
式(51)將
Figure 111103830-A0305-02-0028-187
、r(n)、X(n)分成I與Q通道表示，其中主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數，用矩陣表示成
Figure 111103830-A0305-02-0029-88
，x M (n)、
Figure 111103830-A0305-02-0029-185
、x A (n)、
Figure 111103830-A0305-02-0029-186
用矩陣表示成
Figure 111103830-A0305-02-0029-89
 上式(53)經過微分後如式(54)：
Figure 111103830-A0305-02-0029-90
 拉格朗日乘數λ 1和λ 2表示為式(55)：
Figure 111103830-A0305-02-0029-91
 簡化為式(56)：
Figure 111103830-A0305-02-0029-92
 其中
Figure 111103830-A0305-02-0029-94
 The amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal are estimated by using the proportional normalized least mean method. Compared with NLMS, when the reflected wave path is sparse, PNLMS has very fast initial convergence and tracking. If the impulse response is more dispersed, the convergence speed of PNLMS is not as good as that of NLMS. From the above, it is known that PNLMS is easily affected by the environment. The cost function of the proportional normalized least mean method is expressed as formula (53):
Figure 111103830-A0305-02-0028-87
 Equation (51) will
Figure 111103830-A0305-02-0028-187
, r ( n ), X( n ) are divided into I and Q channels, where the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the main channel signal and the phase channel of the auxiliary channel signal parameters, expressed as a matrix
Figure 111103830-A0305-02-0029-88
, x M ( n ),
Figure 111103830-A0305-02-0029-185
, x A ( n ),
Figure 111103830-A0305-02-0029-186
expressed as a matrix
Figure 111103830-A0305-02-0029-89
 The above formula (53) is as formula (54) after differentiation:
Figure 111103830-A0305-02-0029-90
 The Lagrangian multipliers λ 1 and λ 2 are expressed as formula (55):
Figure 111103830-A0305-02-0029-91
 Simplified to formula (56):
Figure 111103830-A0305-02-0029-92
 in
Figure 111103830-A0305-02-0029-94

P(n)=I-G(n-1)X(n)[X T (n)G(n-1)X(n)]-1 X T (n) 其中誤差值e[n]及主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數表示為式(57)：

Figure 111103830-A0305-02-0030-95
P ( n )=I- G ( n -1) X ( n )[ X T ( n ) G ( n -1) X ( n )] -1 X T ( n ) where the error value e [ n ] and the main The amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal are expressed as formula (57):
Figure 111103830-A0305-02-0030-95

上述步驟S14通過最小平方法、最小均方法、遞迴最小平方法、歸一化最小均方法或比例歸一化最小均方法估測出主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數進行估測後，即可提供主路濾波器20及輔路濾波器22使用，濾除主路訊號與輔路訊號中的IQ不平衡及通道濾波響應。 The above step S14 estimates the amplitude channel filter response parameters of the amplitude channels of the main channel signal and the auxiliary channel signal by the least square method, the least mean method, the recursive least square method, the normalized least mean method or the proportional normalized least mean method , and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal are estimated, and can be used for the main channel filter 20 and the auxiliary channel filter 22 to filter the IQ imbalance between the main channel signal and the auxiliary channel signal and channel filter response.

接著估測預補償參數。參步驟S16，根據主路訊號之通道濾波響應參數及輔路訊號之通道濾波響應參數，補償射頻不完美聯合訊號模型之通道濾波響應，以估算測出預補償參數。在本實施例中係對主路訊號與輔路訊號的振幅通道的振幅預補償參數，以及相位通道的相位預補償參數進行估測。參數估測器14在估算預補償參數時，會加入振幅預補償代數及相位預補償代數至射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型。其中理想射頻聯合訊號模型表示為，式(58)：

Figure 111103830-A0305-02-0030-287
其中
Figure 111103830-A0305-02-0030-288
 其中h M,±(n)為主路訊號之通道濾波響應參數，h A,±(n)為輔路訊號之通道濾波響應參數，w 1(n)為振幅預補償代數，w 2(n)相位預補償代數，R(n)為完美輸出訊號。 The precompensation parameters are then estimated. Referring to step S16, according to the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, the channel filter response of the RF imperfect joint signal model is compensated to estimate the pre-compensation parameters. In this embodiment, the amplitude precompensation parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase precompensation parameters of the phase channel are estimated. When estimating the precompensation parameters, the parameter estimator 14 will add amplitude precompensation algebra and phase precompensation algebra to the radio frequency imperfect joint signal model to establish an ideal radio frequency joint signal model. Among them, the ideal radio frequency joint signal model is expressed as formula (58):
Figure 111103830-A0305-02-0030-287
 in
Figure 111103830-A0305-02-0030-288
 Among them, h M , ± ( n ) is the channel filter response parameter of the main channel signal, h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal, w 1 ( n ) is the amplitude precompensation algebra, w 2 ( n ) Phase precompensation algebra, R ( n ) is the perfect output signal.

接著帶入上述步驟S14估測出的振幅通道濾波響應參數、相位通道濾波響應參數、振幅不平衡參數及相位不平衡參數至理想射頻聯合訊號模型中，以估測出振幅預補償參數及相位預補償參數。詳細來說，將理想射頻聯合訊號模型轉換為矩陣，其表示為式(59)：

Figure 111103830-A0305-02-0031-97
其中H A,+與H A,-為h A,+(n)與h A,-(n)之托普利茲矩陣表示，w 1為振幅預補償參數，w 2為相位預補償參數，由3.1至3.5小節已知上式之h M 與H A ，欲知w1與w2採用Least Square方法使用偽逆矩陣(Pseudoinverse)，求得(60)式。 Then bring the amplitude channel filter response parameters, phase channel filter response parameters, amplitude imbalance parameters and phase imbalance parameters estimated in the above step S14 into the ideal radio frequency joint signal model to estimate the amplitude pre-compensation parameters and phase pre-compensation parameters. compensation parameters. In detail, the ideal RF joint signal model is transformed into a matrix, which is expressed as equation (59):
Figure 111103830-A0305-02-0031-97
 Among them, H A ,+ and H A ,- are represented by the Toeplitz matrix of h A ,+ ( n ) and h A ,- ( n ), w 1 is the amplitude pre-compensation parameter, w 2 is the phase pre-compensation parameter, given by From Sections 3.1 to 3.5, the h M and H A of the above formula are known. If you want to know w 1 and w 2 , use the Least Square method and use the pseudoinverse matrix (Pseudoinverse) to obtain the formula (60).

Figure 111103830-A0305-02-0031-98
Figure 111103830-A0305-02-0031-98

最後進入步驟S18，利用振幅預補償參數及相位預補償參數，補償不完美輸出訊號。透過上述方法即可得到主路訊號與輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及主路訊號與輔路訊號的振幅通道的振幅通道濾波響應參數

Figure 111103830-A0305-02-0031-177
、
Figure 111103830-A0305-02-0031-178
，以及主路訊號與輔路訊號的相位通道的相位通道濾波響應參數
Figure 111103830-A0305-02-0031-179
、
Figure 111103830-A0305-02-0031-180
進行估測，以及振幅預補償參數w 1、相位預補償參數w 2，故本實施例可將上述補償參數提供給能消除射頻洩漏之收發機，以有效補償不完美輸出訊號。 Finally, enter step S18, using amplitude pre-compensation parameters and phase pre-compensation parameters to compensate the imperfect output signal. Through the above method, the amplitude (In-phase, I) unbalance parameter and phase (Quadrature, Q) unbalance parameter of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response of the amplitude channel of the main channel signal and the auxiliary channel signal can be obtained parameter
Figure 111103830-A0305-02-0031-177
,
Figure 111103830-A0305-02-0031-178
, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal
Figure 111103830-A0305-02-0031-179
,
Figure 111103830-A0305-02-0031-180
Estimated, amplitude pre-compensation parameter w 1 , phase pre-compensation parameter w 2 , so this embodiment can provide the above compensation parameters to the transceiver capable of eliminating radio frequency leakage, so as to effectively compensate the imperfect output signal.

在補償不完美輸出訊號時，主路訊號發射器10發射主路訊號x M (n)經過主路濾波器20，以濾除主路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及主路訊號的振幅通道的振幅通道濾波響應參數，及主路訊號的相位通道的相位通道濾波響應參數。同時輔路訊號發射器12發射輔路訊號x A (n)，此時切換器18切換輔路訊號x A (n)進入線性濾波器16，以透過線性濾波器16的振幅預補償參數w 1、相位預補償參數w 2來補償輔路訊號，其再進入到輔路濾波器22，以濾除輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及輔路訊號的振幅通道的振幅通道濾波響應參數，及輔路訊號的相位通道的相位通道濾波響應參數。因此帶有預補償參數的輔路訊號就能完全消除部分洩漏的主路訊號，達到主動射頻洩漏消除之技術。 When compensating the imperfect output signal, the main channel signal transmitter 10 transmits the main channel signal x M ( n ) through the main channel filter 20 to filter out the amplitude (In-phase, I) unbalance parameter and phase of the main channel signal (Quadrature, Q) unbalance parameter, and the amplitude channel filter response parameter of the amplitude channel of the main channel signal, and the phase channel filter response parameter of the phase channel of the main channel signal. At the same time, the auxiliary channel signal transmitter 12 transmits the auxiliary channel signal x A ( n ), at this time, the switcher 18 switches the auxiliary channel signal x A ( n ) into the linear filter 16 to pass through the amplitude pre-compensation parameter w 1 and phase pre-compensation parameter w 1 of the linear filter 16. The compensation parameter w 2 is used to compensate the auxiliary road signal, which then enters the auxiliary road filter 22 to filter out the amplitude (In-phase, I) imbalance parameter and phase (Quadrature, Q) imbalance parameter of the auxiliary road signal, and the auxiliary road signal The amplitude channel filter response parameter of the amplitude channel, and the phase channel filter response parameter of the phase channel of the auxiliary channel signal. Therefore, the auxiliary channel signal with pre-compensation parameters can completely eliminate part of the leaked main channel signal, achieving the technology of active radio frequency leakage elimination.

請參照第四圖，其為經本實施例所應用之方法所產生的實驗數據比較頻譜圖，藉由10兆赫(MHz)的單載波訊號可從頻譜圖上直接觀測本實施例之方法所估測預振幅預補償參數w 1、相位預補償參數w 2是否有效抑制洩漏的主路訊號，且可透過計算執行主動射頻消除前後強度比值。由第四圖能明顯看出預補償前洩漏的主路訊號明顯突出，但預補償後已完全被削除。 Please refer to the fourth figure, which is a comparison spectrogram of experimental data generated by the method applied in this embodiment. The 10 megahertz (MHz) single carrier signal can be directly observed from the spectrogram and estimated by the method of this embodiment Whether the pre-amplitude pre-compensation parameter w 1 and the phase pre-compensation parameter w 2 can effectively suppress the leaked main channel signal can be calculated by calculating the intensity ratio before and after active radio frequency cancellation. From Figure 4, it can be clearly seen that the leaked main channel signal is prominent before pre-compensation, but it has been completely eliminated after pre-compensation.

綜上所述，本發明能估測出訊號中的通道濾波響應以及預補償參數，以解決收發機訊號洩漏的問題，提升收發機間的隔離度性能，提升訊號傳遞效益。同時能估測出訊號中的IQ不平衡的問題，以補償訊號，提升訊號傳遞效益。 To sum up, the present invention can estimate the channel filter response and pre-compensation parameters in the signal, so as to solve the problem of transceiver signal leakage, improve the isolation performance between transceivers, and improve the efficiency of signal transmission. At the same time, it can estimate the IQ imbalance problem in the signal to compensate the signal and improve the efficiency of signal transmission.

唯以上所述者，僅為本發明之較佳實施例而已，並非用來限定本發明實施之範圍。故即凡依本發明申請範圍所述之特徵及精神所為之均等變化或修飾，均應包括於本發明之申請專利範圍內。 The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Therefore, all equivalent changes or modifications based on the features and spirit described in the scope of the application of the present invention shall be included in the scope of the patent application of the present invention.

S10~S18:步驟 S10~S18: steps

Claims (20)

一種全雙工無線接收機之射頻消除自我干擾方法，包括下列步驟：建立一射頻不完美聯合訊號模型；輸入主路訊號與輔路訊號至該射頻不完美聯合訊號模型中，以將該主路訊號與該輔路訊號合成，產生不完美輸出訊號；根據該射頻不完美聯合訊號模型及該不完美輸出訊號，利用最小平方法或最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測，以取得該主路訊號之通道濾波響應參數及該輔路訊號之通道濾波響應參數；根據該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數，補償該射頻不完美聯合訊號模型之該通道濾波響應，以估算測出預補償參數；以及利用該預補償參數補償該不完美輸出訊號。 A radio frequency self-interference elimination method for a full-duplex wireless receiver, comprising the following steps: establishing a radio frequency imperfect combined signal model; inputting a main channel signal and an auxiliary channel signal into the radio frequency imperfect combined signal model, so that the main channel signal Combined with the auxiliary signal to generate an imperfect output signal; according to the RF imperfect joint signal model and the imperfect output signal, use the least square method or the least average method to estimate the channel filter response of the main signal and the auxiliary signal To obtain the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal; according to the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal, the imperfection of the radio frequency is compensated combining the channel filter response of the signal model to estimate a measured precompensation parameter; and compensating the imperfect output signal using the precompensation parameter. 如請求項1所述之全雙工無線接收機之射頻消除自我干擾方法，其中該射頻不完美聯合訊號模型表示為：
Figure 111103830-A0305-02-0034-175
 其中該x M (n)為該主路訊號，該x A (n)為該輔路訊號，該c M (n)為該主路訊號之該通道濾波響應參數，該c A (n)為該輔路訊號之該通道濾波響應參數，該r(n)為該不完美輸出訊號，
Figure 111103830-A0305-02-0034-176
為迴旋運算處理。 The radio frequency self-interference elimination method for a full-duplex wireless receiver as described in claim 1, wherein the radio frequency imperfect joint signal model is expressed as:
Figure 111103830-A0305-02-0034-175
 Wherein the x M ( n ) is the main channel signal, the x A ( n ) is the auxiliary channel signal, the c M ( n ) is the channel filter response parameter of the main channel signal, and the c A ( n ) is the The channel filter response parameter of the auxiliary channel signal, the r ( n ) is the imperfect output signal,
Figure 111103830-A0305-02-0034-176
It is processed by convolution operation. 如請求項2所述之全雙工無線接收機之射頻消除自我干擾方法，其中利用該最小平方法或該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，使用該最小平方法對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號轉換以矩陣表示；及 利用最小平方法的代價函數式，估測該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 2, wherein the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal by using the least square method or the least average method , using the least squares method to estimate the channel filter response of the main channel signal and the auxiliary channel signal, comprising the following steps: converting the imperfect output signal into a matrix representation; and Estimate the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal by using the cost function formula of the least square method. 如請求項2所述之全雙工無線接收機之射頻消除自我干擾方法，其中利用該最小平方法或該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，係使用該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號帶入最小均方法式，表示為：J LMS =avg{e[n]e *[n]}其中
Figure 111103830-A0305-02-0035-99
 該
Figure 111103830-A0305-02-0035-170
為該主路訊號之該通道濾波響應參數及該
Figure 111103830-A0305-02-0035-171
為該輔路訊號之該通道濾波響應參數；將該
Figure 111103830-A0305-02-0035-173
及該
Figure 111103830-A0305-02-0035-174
帶入遞迴關係式，表示為：
Figure 111103830-A0305-02-0035-138
；及
Figure 111103830-A0305-02-0035-101
 對該遞迴關係式進行微分，以估測出該主路訊號之該通道濾波響應參數，及該輔路訊號之該通道濾波響應參數，該遞迴關係式進行微分表示為：
Figure 111103830-A0305-02-0035-104
 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 2, wherein the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal by using the least square method or the least average method , is to use the least average method to estimate the channel filter response of the main channel signal and the auxiliary channel signal, including the following steps: bring the imperfect output signal into the least average method, expressed as: J LMS = avg { e [ n ] e * [ n ]} where
Figure 111103830-A0305-02-0035-99
 Should
Figure 111103830-A0305-02-0035-170
is the channel filter response parameter of the main channel signal and the
Figure 111103830-A0305-02-0035-171
is the channel filter response parameter of the auxiliary channel signal; the
Figure 111103830-A0305-02-0035-173
and the
Figure 111103830-A0305-02-0035-174
Bringing in the recursive relation, expressed as:
Figure 111103830-A0305-02-0035-138
;and
Figure 111103830-A0305-02-0035-101
 Differentiate the recursive relational expression to estimate the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal. The differential expression of the recursive relational expression is expressed as:
Figure 111103830-A0305-02-0035-104
 如請求項2所述之全雙工無線接收機之射頻消除自我干擾方法，其中利用該最小平方法或該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，係使用遞迴最小平方法(recursive least squares， RLS)對該主路訊號與該輔路訊號之該通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號帶入遞迴最小平方法式，以求得該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 2, wherein the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal by using the least square method or the least average method , using the recursive least squares method (recursive least squares, RLS) estimates the channel filter response of the main channel signal and the auxiliary channel signal, including the following steps: bringing the imperfect output signal into a recursive least square method to obtain the channel filter response of the main channel signal parameter and the channel filter response parameter of the auxiliary channel signal. 如請求項2所述之全雙工無線接收機之射頻消除自我干擾方法，其中利用該最小平方法或該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，係使用歸一化最小均方法(Normalized Least Mean Squares，NLMS)對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號帶入歸一化最小均方式，以求得該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 2, wherein the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal by using the least square method or the least average method , is to use the normalized least mean method (Normalized Least Mean Squares, NLMS) to estimate the channel filter response of the main channel signal and the auxiliary channel signal, including the following steps: bring the imperfect output signal into the normalized minimum average method to obtain the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal. 如請求項2所述之全雙工無線接收機之射頻消除自我干擾方法，其中利用該最小平方法或該最小均方法對該主路訊號與該輔路訊號之通道濾波響應進行估測之步驟中，係使用比例歸一化最小均方法(PNLMS)對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該主路訊號之該通道濾波響應參數c M (n)，及該輔路訊號之該通道濾波響應參數c A (n)以矩陣形式表示成：
Figure 111103830-A0305-02-0036-105
 將該主路訊號x M (n)及該輔路訊號x A (n)以矩陣形式表示成：X=[x M (n)x A (n)]：將該
Figure 111103830-A0305-02-0036-169
、該不完美輸出訊號r(n)及該X(n)分成振幅(I)通道與相位(Q)通道，表示式為：
Figure 111103830-A0305-02-0037-107
 其中該
Figure 111103830-A0305-02-0037-165
、該
Figure 111103830-A0305-02-0037-166
為該
Figure 111103830-A0305-02-0037-167
的振幅(I)通道與相位(Q)通道，該r I (n)、該r Q 為該r(n)的振幅(I)通道與相位(Q)通道，該X i,I (n)、該X i,Q (n)為該X(n)的振幅(I)通道與相位(Q)通道；對上式該J PNLMS 進行微分，並使用拉格朗日乘數法運算，以估測該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 2, wherein the step of estimating the channel filter response of the main channel signal and the auxiliary channel signal by using the least square method or the least average method , using the proportional normalized least mean method (PNLMS) to estimate the channel filter response of the main channel signal and the auxiliary channel signal, including the following steps: the channel filter response parameter c M ( n ) of the main channel signal , and the channel filter response parameter c A ( n ) of the auxiliary channel signal is expressed in matrix form as:
Figure 111103830-A0305-02-0036-105
 Express the main channel signal x M ( n ) and the auxiliary channel signal x A ( n ) in matrix form: X=[ x M ( n ) x A ( n )]: the
Figure 111103830-A0305-02-0036-169
, the imperfect output signal r ( n ) and the X ( n ) are divided into an amplitude (I) channel and a phase (Q) channel, the expression is:
Figure 111103830-A0305-02-0037-107
 which the
Figure 111103830-A0305-02-0037-165
,Should
Figure 111103830-A0305-02-0037-166
for the
Figure 111103830-A0305-02-0037-167
The amplitude (I) channel and phase (Q) channel of the r I ( n ), the r Q is the amplitude (I) channel and phase (Q) channel of the r ( n ), the X i,I ( n ) , the X i, Q ( n ) is the amplitude (I) channel and phase (Q) channel of the X ( n ); differentiate the J PNLMS of the above formula, and use the Lagrange multiplier method to estimate Measure the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal. 如請求項1所述之全雙工無線接收機之射頻消除自我干擾方法，其中根據該主路訊號之該通道濾波響應參數及該輔路訊號之該通道濾波響應參數，補償該射頻不完美聯合訊號模型之該通道濾波響應，以估算測出預補償參數之步驟更包括：加入預補償代數至該射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型；及帶入該主路訊號之通道濾波響應參數及該輔路訊號之通道濾波響應參數至該理想射頻聯合訊號模型中，以估測出該預補償參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 1, wherein the radio frequency imperfect joint signal is compensated according to the channel filter response parameter of the main channel signal and the channel filter response parameter of the auxiliary channel signal The channel filter response of the model to estimate the measured pre-compensation parameters further includes: adding pre-compensation algebra to the radio frequency imperfect joint signal model to establish an ideal radio frequency joint signal model; and bringing in the channel filter of the main channel signal The response parameter and the channel filter response parameter of the auxiliary channel signal are put into the ideal radio frequency joint signal model to estimate the precompensation parameter. 如請求項8所述之全雙工無線接收機之射頻消除自我干擾方法，其中該理想射頻聯合訊號模型表示為：
Figure 111103830-A0305-02-0037-164
 其中該x M (n)為該主路訊號，該x A (n)為該輔路訊號，該c M (n)為該主路訊號之該通道濾波響應參數，該c A (n)為該輔路訊號之該通道濾波響應參數，該R(n)為該完美輸出訊號，該q A (n)為該預補償代數，
Figure 111103830-A0305-02-0037-168
為迴旋運算處理。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 8, wherein the ideal radio frequency joint signal model is expressed as:
Figure 111103830-A0305-02-0037-164
 Wherein the x M ( n ) is the main channel signal, the x A ( n ) is the auxiliary channel signal, the c M ( n ) is the channel filter response parameter of the main channel signal, and the c A ( n ) is the The channel filter response parameter of the auxiliary channel signal, the R ( n ) is the perfect output signal, the q A ( n ) is the pre-compensation algebra,
Figure 111103830-A0305-02-0037-168
It is processed by convolution operation. 如請求項9所述之全雙工無線接收機之射頻消除自我干擾方法，其中帶入該主路訊號之通道濾波響應參數及該輔路訊號之通道濾波響應參數至該理想射頻聯合訊號模型中，以估測出該預補償參數之步驟更包括：將
Figure 111103830-A0305-02-0038-162
轉換為矩陣形，表示為：c M =-C A q A 該q A 為該q A (n)的向量表示，該C A 為該c A (n)的常對角矩陣(Toeplitz)表示；使用偽逆矩陣(Pseudoinverse)轉換該q A ，以估測出該預補償參數，偽逆矩陣的q A 表示為：q A =-(C A H C A )-1 C A H c M 。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 9, wherein the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal are introduced into the ideal radio frequency combined signal model, The step of estimating the pre-compensation parameter further includes:
Figure 111103830-A0305-02-0038-162
Converted to a matrix form, expressed as: c M =- C A q A, the q A is the vector representation of the q A ( n ), and the C A is the constant diagonal matrix (Toeplitz) representation of the c A ( n ); The q A is converted using a pseudo-inverse matrix (Pseudoinverse) to estimate the pre-compensation parameter, and the q A of the pseudo-inverse matrix is expressed as: q A =-( C A H C A ) -1 C A H c M . 如請求項1所述之全雙工無線接收機之射頻消除自我干擾方法，其中在根據該射頻不完美聯合訊號模型及該不完美輸出訊號，對該主路訊號與該輔路訊號之通道濾波響應進行估測，以取得該主路訊號之通道濾波響應參數及該輔路訊號之通道濾波響應參數之步驟中，更包括對該主路訊號與該輔路訊號的振幅(In-phase，I)不平衡參數及相位(Quadrature，Q)不平衡參數，及該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 1, wherein the channel filter response of the main channel signal and the auxiliary channel signal is based on the radio frequency imperfect combined signal model and the imperfect output signal In the step of estimating to obtain the channel filter response parameters of the main channel signal and the channel filter response parameters of the auxiliary channel signal, the amplitude (In-phase, I) imbalance between the main channel signal and the auxiliary channel signal is further included parameter and phase (Quadrature, Q) unbalance parameter, and the amplitude channel filter response parameter of the amplitude channel of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameter of the phase channel of the main channel signal and the auxiliary channel signal estimate. 如請求項1所述之全雙工無線接收機之射頻消除自我干擾方法，其中該射頻不完美聯合訊號模型表示為：
Figure 111103830-A0305-02-0038-161
 其中
Figure 111103830-A0305-02-0038-108
 該x M (n)為該主路訊號，該x A (n)為該輔路訊號，該r(n)為該不完美輸出訊號，該
Figure 111103830-A0305-02-0038-163
為該主路訊號之該振幅通道濾波響應參數，該α M 為該主路訊號的該振幅 不平衡參數，該θ M 為該主路訊號的該相位不平衡參數，該
Figure 111103830-A0305-02-0039-157
為該主路訊號之該相位通道濾波響應參數，該
Figure 111103830-A0305-02-0039-158
為該輔路訊號之該振幅通道濾波響應參數，該α A 為該輔路訊號的該振幅不平衡參數，該θ A 為該輔路訊號的該相位不平衡參數，該
Figure 111103830-A0305-02-0039-160
為該輔路訊號之該相位通道濾波響應參數，該h M,±(n)為該主路訊號之該通道濾波響應參數，該h A,±(n)為該輔路訊號之該通道濾波響應參數。 The radio frequency self-interference elimination method for a full-duplex wireless receiver as described in claim 1, wherein the radio frequency imperfect joint signal model is expressed as:
Figure 111103830-A0305-02-0038-161
 in
Figure 111103830-A0305-02-0038-108
 The x M ( n ) is the main channel signal, the x A ( n ) is the auxiliary channel signal, the r ( n ) is the imperfect output signal, the
Figure 111103830-A0305-02-0038-163
is the amplitude channel filter response parameter of the main channel signal, the α M is the amplitude unbalance parameter of the main channel signal, the θ M is the phase imbalance parameter of the main channel signal, the
Figure 111103830-A0305-02-0039-157
is the phase channel filter response parameter of the main channel signal, the
Figure 111103830-A0305-02-0039-158
is the amplitude channel filter response parameter of the auxiliary channel signal, the α A is the amplitude imbalance parameter of the auxiliary channel signal, the θ A is the phase imbalance parameter of the auxiliary channel signal, the
Figure 111103830-A0305-02-0039-160
is the phase channel filter response parameter of the auxiliary channel signal, the h M , ± ( n ) is the channel filter response parameter of the main channel signal, and the h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal . 如請求項12所述之全雙工無線接收機之射頻消除自我干擾方法，其中對該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用該最小平方法(Least Square，LS)進行估測，包括下列步驟：將該不完美輸出訊號轉換以矩陣表示；利用最小平方法的代價函數式，估測出該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 12, wherein the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response parameters of the main channel signal and the auxiliary channel signal In the step of estimating the phase channel filter response parameters of the phase channel, the least square method (Least Square, LS) is used for estimation, including the following steps: converting the imperfect output signal into a matrix representation; using the least square method The cost function formula of is used to estimate the amplitude channel filter response parameters of the amplitude channels of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal. 如請求項12所述之全雙工無線接收機之射頻消除自我干擾方法，其中對該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用該最小均方法(Least Mean Square，LMS)進行估測，包括下列步驟：將該不完美輸出訊號帶入最小均方法式，表示為：J LMS =avg{e[n]e *[n]}其中
Figure 111103830-A0305-02-0039-139
 該h M,±(n)為該主路訊號之該通道濾波響應參數，該h A,±(n)為該輔路訊號之該通道濾波響應參數；將該
Figure 111103830-A0305-02-0040-153
、該
Figure 111103830-A0305-02-0040-154
、該
Figure 111103830-A0305-02-0040-155
、該
Figure 111103830-A0305-02-0040-156
帶入遞迴關係式，表示為：
Figure 111103830-A0305-02-0040-110
 對該遞迴關係式進行微分，以估測出該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測，遞迴關係式進行微分表示為：
Figure 111103830-A0305-02-0040-112
 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 12, wherein the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response parameters of the main channel signal and the auxiliary channel signal In the step of estimating the phase channel filter response parameters of the phase channel, the Least Mean Square (LMS) method is used for estimation, including the following steps: bringing the imperfect output signal into the least mean square method, expressing is: J LMS = avg { e [ n ] e * [ n ]} where
Figure 111103830-A0305-02-0039-139
 The h M , ± ( n ) is the channel filter response parameter of the main channel signal, and the h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal;
Figure 111103830-A0305-02-0040-153
,Should
Figure 111103830-A0305-02-0040-154
,Should
Figure 111103830-A0305-02-0040-155
,Should
Figure 111103830-A0305-02-0040-156
Bringing in the recursive relation, expressed as:
Figure 111103830-A0305-02-0040-110
 Differentiate the recursive relation to estimate the amplitude channel filter response parameters of the main channel signal and the amplitude channel of the auxiliary channel signal, and the phase channel filter response parameters of the main channel signal and the phase channel of the auxiliary channel signal. Estimated, the recursive relation is differentiated and expressed as:
Figure 111103830-A0305-02-0040-112
 如請求項12所述之全雙工無線接收機之射頻消除自我干擾方法，其中對該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用遞迴最小平方法(recursive least squares，RLS)對該主路訊號與該輔路訊號之該通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號帶入遞迴最小平方法式，以求得該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 12, wherein the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response parameters of the main channel signal and the auxiliary channel signal In the step of estimating the phase channel filter response parameters of the phase channel, the channel filter response of the main channel signal and the auxiliary channel signal is estimated using the recursive least squares method (recursive least squares, RLS), including the following steps : Bring the imperfect output signal into the recursive least square method to obtain the amplitude channel filter response parameters of the main channel signal and the amplitude channel of the auxiliary channel signal, and the phase of the phase channel of the main channel signal and the auxiliary channel signal Channel filter response parameters are estimated. 如請求項12所述之全雙工無線接收機之射頻消除自我干擾方法，其中對該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用歸一化最小均方法(Normalized Least Mean Squares，NLMS)對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該不完美輸出訊號帶入歸一化最小均方式，以求得該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 12, wherein the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response parameters of the main channel signal and the auxiliary channel signal In the step of estimating the phase channel filter response parameters of the phase channel, the normalized least mean method (Normalized Least Mean Squares, NLMS) is used to estimate the channel filter response of the main channel signal and the auxiliary channel signal, including the following Steps: Bring the imperfect output signal into the normalized minimum average method to obtain the amplitude channel filter response parameters of the main channel signal and the amplitude channel of the auxiliary channel signal, and the phase channel of the main channel signal and the auxiliary channel signal Estimate the response parameters of the phase channel filter. 如請求項12所述之全雙工無線接收機之射頻消除自我干擾方法，其中對該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測之步驟中，係使用比例歸一化最小均方法(PNLMS)對該主路訊號與該輔路訊號之通道濾波響應進行估測，包括下列步驟：將該主路訊號之該通道濾波響應參數h M,±(n)，及該輔路訊號之該通道濾波響應參數h A,±(n)以矩陣形式表示成：
Figure 111103830-A0305-02-0041-113
 將該主路訊號x M (n)、
Figure 111103830-A0305-02-0041-150
及該輔路訊號x A (n)、
Figure 111103830-A0305-02-0041-151
以矩陣形式表示成：
Figure 111103830-A0305-02-0041-114
 將該
Figure 111103830-A0305-02-0041-152
、該不完美輸出訊號r(n)及該X(n)分成振幅(I)通道與相位(Q)通道，表示式為：
Figure 111103830-A0305-02-0042-116
 其中該
Figure 111103830-A0305-02-0042-148
、該
Figure 111103830-A0305-02-0042-149
的振幅(I)通道與相位(Q)通道，該r I (n)、該r Q 為該r(n)的振幅(I)通道與相位(Q)通道，該X i,I (n)、該X i,Q (n)為該X(n)的振幅(I)通道與相位(Q)通道；對上式該J PNLMS 進行微分，並使用拉格朗日乘數法運算，以估測出該主路訊號與該輔路訊號的振幅通道的振幅通道濾波響應參數，以及該主路訊號與該輔路訊號的相位通道的相位通道濾波響應參數進行估測。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 12, wherein the amplitude channel filter response parameters of the amplitude channel of the main channel signal and the auxiliary channel signal, and the amplitude channel filter response parameters of the main channel signal and the auxiliary channel signal In the step of estimating the phase channel filter response parameter of the phase channel, the proportional normalized least mean method (PNLMS) is used to estimate the channel filter response of the main channel signal and the auxiliary channel signal, including the following steps: The channel filter response parameter h M , ± ( n ) of the main channel signal, and the channel filter response parameter h A , ± ( n ) of the auxiliary channel signal are expressed in matrix form:
Figure 111103830-A0305-02-0041-113
 The main road signal x M ( n ),
Figure 111103830-A0305-02-0041-150
and the auxiliary road signal x A ( n ),
Figure 111103830-A0305-02-0041-151
Expressed in matrix form as:
Figure 111103830-A0305-02-0041-114
 will
Figure 111103830-A0305-02-0041-152
, the imperfect output signal r ( n ) and the X ( n ) are divided into an amplitude (I) channel and a phase (Q) channel, the expression is:
Figure 111103830-A0305-02-0042-116
 which the
Figure 111103830-A0305-02-0042-148
,Should
Figure 111103830-A0305-02-0042-149
The amplitude (I) channel and phase (Q) channel of the r I ( n ), the r Q is the amplitude (I) channel and phase (Q) channel of the r ( n ), the X i,I ( n ) , the X i, Q ( n ) is the amplitude (I) channel and phase (Q) channel of the X ( n ); differentiate the J PNLMS of the above formula, and use the Lagrange multiplier method to estimate The amplitude channel filter response parameters of the amplitude channels of the main channel signal and the auxiliary channel signal, and the phase channel filter response parameters of the phase channel of the main channel signal and the auxiliary channel signal are measured for estimation. 如請求項1所述之全雙工無線接收機之射頻消除自我干擾方法，其中根據該通道濾波響應參數，補償該射頻不完美聯合訊號模型之該通道濾波響應，並估算測出預補償參數之步驟更包括對該主路訊號與該輔路訊號的振幅通道的振幅預補償參數，以及相位通道的相位預補償參數進行估測，其步驟更包括：加入該振幅預補償代數及該相位預補償代數至該射頻不完美聯合訊號模型，以建立理想射頻聯合訊號模型；及帶入該振幅通道濾波響應參數、該相位通道濾波響應參數、該振幅不平衡參數及該相位不平衡參數至該理想射頻聯合訊號模型中，以估測出該振幅預補償參數及該相位預補償參數。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 1, wherein the channel filter response of the radio frequency imperfect joint signal model is compensated according to the channel filter response parameter, and the pre-compensation parameter is estimated. The step further includes estimating amplitude precompensation parameters of the amplitude channels of the main channel signal and the auxiliary channel signal, and estimating phase precompensation parameters of the phase channel, and the steps further include: adding the amplitude precompensation algebra and the phase precompensation algebra to the radio frequency imperfect joint signal model to establish an ideal radio frequency joint signal model; and bring the amplitude channel filter response parameter, the phase channel filter response parameter, the amplitude imbalance parameter and the phase imbalance parameter into the ideal radio frequency joint signal model In the signal model, the amplitude precompensation parameter and the phase precompensation parameter are estimated. 如請求項18所述之全雙工無線接收機之射頻消除自我干擾方法，其中該理想射頻聯合訊號模型表示為：其中該理想射頻聯合訊號模型表示為：
Figure 111103830-A0305-02-0042-279
 其中
Figure 111103830-A0305-02-0042-147
Figure 111103830-A0305-02-0043-143
 其中該h M,±(n)為該主路訊號之該通道濾波響應參數，該h A,±(n)為該輔路訊號之該通道濾波響應參數，該w 1(n)為該振幅預補償代數，該w 2(n)該相位預補償代數，該R(n)為該完美輸出訊號，
Figure 111103830-A0305-02-0043-144
為迴旋運算處理。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in Claim 18, wherein the ideal radio frequency combined signal model is expressed as: wherein the ideal radio frequency combined signal model is expressed as:
Figure 111103830-A0305-02-0042-279
 in
Figure 111103830-A0305-02-0042-147
Figure 111103830-A0305-02-0043-143
 Wherein h M , ± ( n ) is the channel filter response parameter of the main channel signal, the h A , ± ( n ) is the channel filter response parameter of the auxiliary channel signal, and w 1 ( n ) is the amplitude preset Compensation algebra, the w 2 ( n ) the phase pre-compensation algebra, the R ( n ) is the perfect output signal,
Figure 111103830-A0305-02-0043-144
It is processed by convolution operation. 如請求項19所述之全雙工無線接收機之射頻消除自我干擾方法，其中帶入振幅通道濾波響應參數、相位通道濾波響應參數、振幅不平衡參數及相位不平衡參數至理想射頻聯合訊號模型中，以估測出振幅預補償參數及相位預補償參數之步驟更包括：將該理想射頻聯合訊號模型轉換為矩陣，其表示為：
Figure 111103830-A0305-02-0043-117
 其中該H A,+與該H A,-為該h A,+(n)與該h A,-(n)之托普利茲(Toeplitz)矩陣表示，該w 1為該振幅預補償參數，該w 2為該相位預補償參數；及採用最小平方法，並使用偽逆矩陣(Pseudoinverse)，求得該w 1及該w 2表示為：
Figure 111103830-A0305-02-0043-140
 其中該H為該H A,+與該H A,-的合成矩陣，該H +=(H H H)-1 H H 。 The radio frequency self-interference elimination method of a full-duplex wireless receiver as described in claim 19, wherein the amplitude channel filter response parameter, the phase channel filter response parameter, the amplitude imbalance parameter and the phase imbalance parameter are introduced into the ideal radio frequency joint signal model In , the step of estimating the amplitude precompensation parameter and the phase precompensation parameter further includes: converting the ideal RF joint signal model into a matrix, which is expressed as:
Figure 111103830-A0305-02-0043-117
 Wherein the H A ,+ and the H A ,- are the Toeplitz matrix representations of the h A ,+ ( n ) and the h A ,- ( n ), the w 1 is the amplitude precompensation parameter, The w 2 is the phase pre-compensation parameter; and the least square method is adopted, and the pseudo inverse matrix (Pseudoinverse) is used to obtain the w 1 and the w 2 are expressed as:
Figure 111103830-A0305-02-0043-140
 Wherein the H is the resultant matrix of the H A ,+ and the H A ,- , the H + =( H H H ) -1 H H .
TW111103830A 2022-01-28 2022-01-28 Radio frequency self-interference elimination method for full-duplex wireless receiver TWI794001B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW111103830A TWI794001B (en) 2022-01-28 2022-01-28 Radio frequency self-interference elimination method for full-duplex wireless receiver

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW111103830A TWI794001B (en) 2022-01-28 2022-01-28 Radio frequency self-interference elimination method for full-duplex wireless receiver

Publications (2)

Publication Number Publication Date
TWI794001B true TWI794001B (en) 2023-02-21
TW202332227A TW202332227A (en) 2023-08-01

Family

ID=86689399

Family Applications (1)

Application Number Title Priority Date Filing Date
TW111103830A TWI794001B (en) 2022-01-28 2022-01-28 Radio frequency self-interference elimination method for full-duplex wireless receiver

Country Status (1)

Country Link
TW (1) TWI794001B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080008258A1 (en) * 2006-07-05 2008-01-10 Yasuhiko Tanabe Wireless receiving apparatus and method
US20090213944A1 (en) * 2008-02-25 2009-08-27 Grant Stephen J Receiver Parametric Covariance Estimation for Precoded MIMO Transmissions
TW201724818A (en) * 2015-12-24 2017-07-01 元智大學 United estimation pre-compensation method for solving imperfection in downstream transmission system capable of transmitting a more perfect signal by compensating for a receiving signal in advance on a transmitting end
US20170353288A1 (en) * 2015-07-16 2017-12-07 LGS Innovations LLC Self-Interference Channel Estimation System and Method
US20200351072A1 (en) * 2012-05-13 2020-11-05 Amir Keyvan Khandani Full duplex wireless transmission with self-interference cancellation

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080008258A1 (en) * 2006-07-05 2008-01-10 Yasuhiko Tanabe Wireless receiving apparatus and method
US20090213944A1 (en) * 2008-02-25 2009-08-27 Grant Stephen J Receiver Parametric Covariance Estimation for Precoded MIMO Transmissions
US20200351072A1 (en) * 2012-05-13 2020-11-05 Amir Keyvan Khandani Full duplex wireless transmission with self-interference cancellation
US20170353288A1 (en) * 2015-07-16 2017-12-07 LGS Innovations LLC Self-Interference Channel Estimation System and Method
TW201724818A (en) * 2015-12-24 2017-07-01 元智大學 United estimation pre-compensation method for solving imperfection in downstream transmission system capable of transmitting a more perfect signal by compensating for a receiving signal in advance on a transmitting end

Also Published As

Publication number Publication date
TW202332227A (en) 2023-08-01

Similar Documents

Publication Publication Date Title
US10574428B2 (en) Self-interference channel estimation system and method
CN109617560B (en) IQ signal calibration compensation method
Korpi et al. Adaptive nonlinear digital self-interference cancellation for mobile inband full-duplex radio: Algorithms and RF measurements
US9270308B2 (en) Adaptive radio-frequency interference cancelling device, method, and receiver
US9698836B2 (en) Systems and methods for mitigation of self-interference in spectrally efficient full duplex communications
JP6100406B2 (en) Multi-tap adaptive filter for transmit signal leakage cancellation
US7043208B2 (en) Method and apparatus to reduce interference in a communication device
CN107026673B (en) Digital-assisted analog domain interference cancellation method and device
Erdem et al. Monostatic antenna in-band full duplex radio: Performance limits and characterization
US10680724B2 (en) Methods and apparatus for duplexing signals
Emara et al. A full duplex transceiver with reduced hardware complexity
TWI677202B (en) Wireless transceiver capable of offsetting internal signal leakage
CN110830076B (en) Wireless transceiver capable of canceling internal signal leakage
TWI794001B (en) Radio frequency self-interference elimination method for full-duplex wireless receiver
Ge et al. Dual-stage co-site RF interference canceller for wideband direct-conversion receivers using reduced observation chain
Jing et al. Implementation of time-window based analog self-interference cancellation for full-duplex radios
TWI769593B (en) Active RF Leakage Elimination Method for Improving Transceiver Isolation
Tapio et al. Analog self-interference cancellation with automatic gain control for full-duplex transceivers
CN109921810B (en) Self-interference suppression method and device suitable for radio frequency domain
Rani et al. Digital Self-Interference Cancellation: Algorithmic Performance Analysis and Software Defined Radio Based Realization
KR101740922B1 (en) Digital Apparatus and Method for Cancelling Transmitter Leakage Signal
Luo et al. A Novel Blind Compensation Scheme for I/Q Imbalance of Wideband Zero-IF Transmitter
Kim et al. Real-Time Implementation of Optimal Nonlinear Self-Interference Cancellation for Full-Duplex Radio
EP2464072B1 (en) Frequency reuse in wireless point-to-point communication systems
JPH03129928A (en) Echo canceller for automobile telephone modem