KR102102059B1 - Full-duplex communication equipment for cancelling self-interference and method for cancelling self-interference - Google Patents

Abstract

According to the present invention, a plurality of antennas and power amplifiers are divided into a first mode and a second mode according to a section in which linear and nonlinear characteristics are operated to output amplified transmission signals to a plurality of antennas. Pre-compensate the characteristics at the digital signal level according to the pre-compensation function to receive the transmission symbol that digitally converts the received signal received through the multiple transmitters and multiple antennas and the transmission symbol that digitally modulates the transmission data from the transmitter. Received, estimates the first self-interference channel gain using the received and transmitted symbols based on the linearity of the self-interference signal in the first mode, obtains the first reconstructed self-interference signal, and is pre-compensated and transmitted in the second mode From the received symbol and the transmitted symbol and the first self-interference channel gain, the second self-interference channel gain and the preceding compensation coefficient It is possible to provide a full-duplex communication terminal and a method for canceling self-interference, including a plurality of receivers that alternately estimate iteratively and obtain the preceding compensation function according to the preceding compensation coefficient and transmit it to the transmitter.

Description

A full-duplex communication terminal capable of removing self-interference and a method for removing self-interference thereof {FULL-DUPLEX COMMUNICATION EQUIPMENT FOR CANCELLING SELF-INTERFERENCE AND METHOD FOR CANCELLING SELF-INTERFERENCE}

The present invention relates to a full-duplex communication terminal capable of removing self-interference and a method for removing self-interference thereof, and to a full-duplex communication terminal capable of digitally removing self-interference and a method for removing self-interference.

Recently, as demand for frequency resources increases exponentially, research has been conducted to efficiently use limited frequency resources. Current wireless communication systems use a half-duplex method such as time division duplex (TDD) or frequency division duplex (FDD) using time or frequency orthogonality.

However, the full-duplex method transmits and receives data at the same time and frequency at one node. Therefore, using the full-duplex communication method, it is possible to obtain twice the frequency efficiency compared to half-duplex communication.

Since full-duplex communication transmits and receives at the same time and frequency, a phenomenon occurs when riding through its receiver during transmission. This signal is referred to as a self-interference signal, and in order to realize full-duplex communication, the self-interference signal must be removed to be below a predetermined noise level.

However, the self-interference signal received at the receiver is digitized by an analog-to-digital converter (ADC) and generates a quantization error. Therefore, even if the receiver performs the digital signal processing with high precision, it cannot remove the self-interference signal beyond the dynamic range of the ADC. That is, in order to completely cancel the self-interference signal, the receiver must simultaneously and simultaneously remove the self-interference signal in a digital method and an analog method.

In addition, an inter-modulation distortion (IMD) component occurs in a self-interference signal through various RF hardware in a transceiver. In particular, the nonlinear component generated in the power amplifier of the transmitter is an important factor that makes it difficult for the receiver to remove the interference signal. However, early self-interference cancellation studies did not take into account the non-linear components occurring in the transceiver and were implemented with a simple algorithm using a linear model, so it was not possible to accurately remove self-interference.

Although in recent years, some studies are underway to remove self-interference in consideration of non-linear components, there is a limitation that it is difficult to practically apply due to the high computational complexity for removing non-linear components.

Korean Open Patent No. 10-2017-0061087 (released on June 2, 2017)

An object of the present invention is to provide a full-duplex communication terminal capable of digitally removing self-interference and a method for removing self-interference.

Another object of the present invention is to provide a full-duplex communication terminal that can be manufactured in a compact size and a method for removing self-interference thereof.

Full-duplex communication terminal according to an embodiment of the present invention for achieving the above object is a plurality of antennas; According to the section in which the power amplifier operates with linear and nonlinear characteristics, the amplified transmission signal is outputted to the plurality of antennas by dividing into a first mode and a second mode, and the nonlinear characteristics of the power amplifier are compensated in advance in the second mode. A plurality of transmitters for outputting the transmission signal by precompensating at a digital signal level according to a function; And a reception symbol obtained by digitally converting a received signal received through the plurality of antennas and a transmission symbol obtained by digitally modulating transmission data from the transmitter, and based on the linearity of the self-interference signal in the first mode. The first self-interference channel gain is estimated by using the transmission symbol, a first reconstructed self-interference signal is obtained, and the received symbol, the transmission symbol, and the first self-interference channel gain are compensated in advance in the second mode. A plurality of receivers for alternately repeatedly estimating a second self-interference channel gain and a preceding compensation coefficient, and acquiring and transmitting the preceding compensation function according to the preceding compensation coefficient to the transmitter; It includes.

The receiver includes a down-converter for down-converting the received signal into a baseband signal; An ADC which converts the down-converted received signal into a digital signal and outputs the received symbol; In the first mode, the first self-interference channel gain is repeatedly estimated according to an RLS algorithm so that the difference between the received symbol and the first reconstructed self-interference signal is minimized, and the pre-compensated reception according to the preceding compensation function in the second mode The second self-interference channel gain is estimated to minimize the difference between the symbol and the second reconstructed self-interference signal, and between the second reconstructed self-interference signal obtained from the estimated second self-interference channel gain and the received symbol and the preceding compensation coefficient A compensation function obtaining unit estimating the preceding compensation coefficient according to a predetermined relationship, obtaining an updated preceding compensation function according to the estimated preceding compensation coefficient, and transmitting the updated compensation function to the transmitter; It may include.

In the second mode, the compensation function acquiring unit is a self-interference cancellation signal (y _i.DC [n] or an error signal (e) that is a difference between the received symbol and the second reconstructed self-interference signal when estimating the preceding compensation coefficient based on the RLS algorithm in the second mode. _i [n])) is estimated by applying the forgetting factor to adjust the reflection ratio, and the forgetting factor is adjusted based on the Mean Square Error (MSE) method according to the signal level of the desire signal (s _i [n]) Can be.

The self-interference cancellation method of a full-duplex communication terminal according to another embodiment of the present invention for achieving the above object transmits a transmission signal in a section in which the power amplifier of the transmitter operates in a linear characteristic in the first mode, and the transmission data is transmitted from the transmitter. Obtaining a digitally modulated transmission symbol and a reception symbol obtained by digitally converting a received signal; Estimating a first self-interference channel gain and obtaining a first reconstructed self-interference signal based on the linearity of the self-interference signal; In the second mode, the power amplifier is pre-compensated for the non-linear characteristic at a digital signal level according to a pre-compensation function in a period in which the power amplifier operates with a non-linear characteristic, and transmits a transmission signal. Obtaining a digitally converted reception symbol; Estimating a second self-interference channel gain and a pre-compensation coefficient by alternately iterating from the pre-compensated transmitted symbol and the transmission symbol and the first self-interference channel gain; And obtaining the preceding compensation function according to the preceding compensation coefficient and transmitting it to the transmitter. It includes.

Therefore, the full-duplex communication terminal capable of removing self-interference and the self-interference cancellation method according to an embodiment of the present invention enable the self-interference signal to be digitally removed. Therefore, the transceiver of the full-duplex communication terminal can be manufactured in a small size. Also, the nonlinear component of the self-interference signal can be easily removed by using the RLS algorithm.

1 shows the structure of a full-duplex communication terminal.
2 shows a configuration of a terminal according to an embodiment of the present invention.
3 shows a detailed configuration of the compensation function obtaining unit of FIG. 2.
4 shows a detailed configuration of the preceding compensation coefficient estimator of FIG. 3.
5 is a view for explaining the concept of setting a standard for optimizing the forgetting factor.
6 shows a self-interference cancellation method of a full-duplex communication terminal according to an embodiment of the present invention.
7 shows a simulation result of self-interference cancellation performance of a full-duplex communication terminal according to an embodiment of the present invention.

In order to fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings and the contents described in the accompanying drawings, which illustrate preferred embodiments of the present invention.

Hereinafter, the present invention will be described in detail by explaining preferred embodiments of the present invention with reference to the accompanying drawings. However, the present invention may be implemented in various different forms, and is not limited to the described embodiments. In addition, in order to clearly describe the present invention, parts irrelevant to the description are omitted, and the same reference numerals in the drawings indicate the same members.

Throughout the specification, when a part “includes” a certain component, this means that other components may be further included, rather than excluding other components, unless specifically stated to the contrary. In addition, terms such as "... unit", "... group", "module", and "block" described in the specification mean a unit that processes at least one function or operation, which is hardware or software or hardware. And software.

1 shows the structure of a full-duplex communication terminal.

1, a full-duplex communication terminal includes a plurality of base stations (BS1, BS2) and a plurality of user terminals (UE). In addition, in the full-duplex communication terminal, each of the plurality of base stations BS1 and BS2 and the plurality of user terminals UE transmits and receives signals with other base stations or other user terminals. At this time, a plurality of base stations (BS1, BS2) and a plurality of user terminals (UE) can simultaneously transmit and receive, including a transmitter and a receiver.

Since a plurality of base stations BS1 and BS2 and a plurality of user terminals UE can simultaneously transmit and receive, a self-interference signal is generated in which the transmitted signal is received again during transmission. The self-interference signal is largely generated in a base station or a user terminal including a plurality of antennas supporting a Multiple Input Multiple Output (MIMO) scheme. This self-interference signal should be removed because it cannot act as a noise and thus cannot determine the signal to be normally received from another base station (BS1, BS2) or a user terminal (UE).

The self-interference signal is a signal that the transmission signal transmitted from the base station (BS1, BS2) or the user terminal (UE) is received again by the base station, and the base stations (BS1, BS2) and the user terminal (UE) transmit the signal that is transmitted intrinsically. Know. However, as described above, the signals to be transmitted by the base stations BS1 and BS2 or the user terminal UE are transmitted by generating intermodulation distortion components through various RF hardware. In addition, the transmitted signal includes a self-interference distortion component by a self-interference channel between multiple antennas and is received by itself.

Due to these distortion components, the base stations BS1 and BS2 and the user terminal UE are difficult to accurately understand the self-interference signal, and as a result, it is not easy to remove the self-interference signal.

Therefore, in order to remove the self-interference signal as much as possible, it is necessary to be able to check both the intermodulation distortion component and the channel distortion component for the signals transmitted by the base stations BS1 and BS2 and the user terminal UE.

2 shows a configuration of a terminal according to an embodiment of the present invention.

In FIG. 2, the terminal may be a base station (BS1, BS2) or a user terminal (UE) of FIG. 1, and the terminal includes a plurality of transmitters 100, a plurality of receivers 200, and a plurality of antennas 300. Hereinafter, for convenience of description, the configuration and operation of the i-th (where i is a natural number) transmitter 100 and the i-th receiver 200 among the plurality of transmitters 100 and the plurality of receivers 200 will be described.

Each of the plurality of transmitters 100 includes a coding and modulation unit 110, a preceding compensation unit 120, a DAC 130, an up converter 140 and a power amplifier 150.

In the i-th transmitter 100, the coding and modulating unit 110 receives transmission data (x _i [n]) to be transmitted and codes and digitally modulates it according to a predetermined communication method to transmit symbols (X _i [n]) Output Here, the transmission data (x _i [n]) may be applied from a signal processing unit (not shown).

In addition, the preceding compensation unit 120 is coded and modulated by the coding and modulation unit 110 by using the preceding compensation function _Pi (·) applied from the compensation function acquisition unit 240 of the receiver 200 and applied Pre-compensation for the transmitted symbol is performed. The pre-compensation unit 120 precompensates for the nonlinearity of the self-interference signal received by the receiver 200 in the transmission symbol according to the pre-compensation function _Pi (·), and thereby the self-interference signal received by the receiver 200 Let it have linearity.

The nonlinearity of the self-interference signal received by the receiver 200 is mostly caused by the nonlinear characteristics of the power amplifier 150 of the transmitter 100. Therefore, it is very important for the preceding compensator 120 to compensate for the nonlinear characteristics of the power amplifier 150, and when compensating for the nonlinear characteristics of the power amplifier 150, a plurality of transmitting antennas 310 and a plurality of receiving terminals The compensation should be performed considering that the gain h _i of the self-interference channel SI _i between the antennas 320 may be varied together.

Since each of the plurality of transmitters 100 precompensates and transmits the nonlinearity of the self-interference signal received by the receiver 200 in each of the receivers 200 in advance, each of the plurality of receivers 200 has a linear magnetic field from the received signal. Interference signals can be very easily removed.

The DAC 130 receives the pre-compensation signal, which is the pre-compensated digital signal, from the pre-compensation unit 120 and converts it into an analog signal. In addition, the up-converter 140 synthesizes a baseband signal converted from the DAC 130 into an analog signal with a carrier having a predetermined frequency and modulates the signal upward.

The power amplifier 150 receives and amplifies the up-modulated signal from the up-converter 140 to amplify and output a transmission signal (x _i.PA [n]) to a corresponding transmission antenna 310 among a plurality of transmission antennas 310. do. Since the power amplifier 150 has a non-linear characteristic, the transmission signal (x _i.PA [n]) may be non-linearly distorted and output. However, in this case, since the preceding compensation unit 120 has previously compensated for the nonlinear characteristic, the receiver 200 may receive a self-interference signal in which the nonlinear characteristic of the power amplifier 150 is compensated.

Each of the plurality of transmission antennas 310 receives and transmits a transmission signal (x _i.PA [n]) from a corresponding transmitter 100 among the plurality of transmitters 100. At this time, the radiated signal may be received by a plurality of receiving antennas 320. That is, a self-interference signal may be received through a magnetic indirect channel (SI) between a plurality of transmitting antennas 310 and a plurality of receiving antennas 320 of the terminal. Along with this, a desired signal (s _i [n]) to be received from another terminal may be received by the plurality of receiving antennas 320. In addition, an inter-cell interference signal (p _i [n]), which is a transmission / reception signal between different terminals, may be received together.

Meanwhile, each of the plurality of receivers 200 receives a received signal received by the corresponding receiving antenna 320 among the plurality of receiving antennas 320.

Here, for convenience of description, the plurality of transmit antennas 310 and the plurality of receive antennas 320 are illustrated separately, but the plurality of transmit antennas 310 and the plurality of receive antennas 320 are not distinguished and the transmit / receive antennas ( 300).

Each of the plurality of receivers 200 includes a low noise amplifier 210, a down converter 220, an ADC 230, a compensation function acquisition unit 240, a self-interference signal removal unit 250, and a decoding and demodulation unit 260. Includes.

In the i-th receiver 200, the low-noise amplifier 210 amplifies the received signal transmitted from the corresponding reception antenna 320, and the down converter 220 synthesizes the signal applied from the low-noise amplifier 210 with a carrier wave. Down-modulation is performed to convert to a baseband signal. Then, the ADC 230 converts the down-modulated baseband analog signal to digital to obtain a received symbol (y _i.res [n]).

The compensation function obtaining unit 240 receives the received symbol y _i.res [n] from the ADC 230 to estimate the gain h _i of the self-interference channel SI _i , and the estimated self-interference channel gain The self-interference signal is reconstructed using (h _i ) and a transmission symbol transmitted from the transmitter 100. And rebuilding a magnetic interference signal and the received symbol (y _i.res [n]) estimating a prior compensation coefficient (C _i) from, and the estimated prior compensation coefficient (C _i) in accordance with prior compensation function (P _i (· )) Is obtained and transmitted to the preceding compensation unit 120 of the corresponding transmitter 100. This is to allow the preceding compensator 120 to compensate for the nonlinear characteristics of the self-interference signal as described above.

Also obtained compensation function unit 240 is received symbols (y _i.res [n]) by subtracting the reconstructed self-interference signal, the interference cancellation signal (y _i.DS [n]) to obtain a decoding and demodulation unit in the (250).

At this time, the compensation function obtaining unit 240 may estimate the self-interference channel gain (h _i ) and the preceding compensation coefficient (C _i ) in a state in which the preceding compensation unit 120 does not normally compensate for the mutual transmission symbol, The self-interference signal can be reconstructed.

Accordingly, the compensation function acquiring unit 240 of the present embodiment may be divided into a first mode (mode1) in which the power amplifier PA operates with a linear characteristic and a second mode (mode1) operating in a nonlinear characteristic. .

The detailed operation of the compensation function acquisition unit 240 will be described later.

The decoding and demodulation unit 250 demodulates and decodes the interference cancellation signal y _i.DS [n] applied from the compensation function acquiring unit 240 in a predetermined manner and outputs the received data y _i [n]. do.

In the present embodiment, the pre-compensation unit 120 implemented as a digital circuit by allowing the pre-compensation unit 120 of each of the plurality of transmitters 100 to receive the pre-compensation function _Pi (·) from the receiver 200 The self-interference signal can be easily removed with the and compensation function acquisition unit 240 only.

When the compensation function acquiring unit 240 estimates the preceding compensation coefficient C _i again, the compensation function acquiring unit 240 first performs the preceding compensation by the preceding compensation unit 120 of the transmitter 100. When the transmission signal (x _i.PA [n]) is transmitted without being performed, the self-interference signal to be removed by the receiver 200 is checked. That is, the self-interference signal for the received down-converted symbol is analyzed by the i-th receiver 200 in a state in which no preceding compensation is performed.

When the digitally converted n-th reception symbol is called a reception symbol (y _i.res [n]), a reception symbol (y _i.res [n]) including a self-interference signal is expressed by Equation (1).

Where h _ij and M _ij represent the channel gain and length of the self-interference channel (SI) between the i th transmit antenna and the j th receive antenna, respectively, and b _j ^(k) is the k th nonlinear coefficient of the i th antenna, φ _k (x _j ) is the basic function of the system model with self-interference φ _k (x _j ) = | x _j [n] | ^k-1 x _j [n] (where x _j [n] is the transmission data of the j th transmitter), N _t is the number of antennas, K _PA is the length of the transmitter's power amplifier 150, and s _i [n] is required The signal p _i [n] denotes an inter-cell interference signal, and w _i denotes thermal noise.

That is, the received signal (y _i.res [n]) includes not only the self-interference signal (h _ij [m] b _j ^(k) φ _k (x _j ) [nm]), but also the required signal (s _i [n]). The inter-cell interference signal p _i [n] and the thermal noise w _i may be included together. In general, the required signal s _i [n] is dominant compared to the inter-cell interference signal p _i [n] and the thermal noise w _i . That is, it is not only often less than a predetermined noise level, but various methods for removing it have been proposed. Therefore, it is not considered here that the inter-cell interference signal p _i [n] and the thermal noise w _i are less than the noise level.

However, in the case of a self-interference signal, it is not only generated in the full-duplex communication terminal, but is a signal transmitted from the adjacent antenna 310 and must be removed because it is applied at a relatively large power level than the required signal s _i [n].

On the other hand, when the prior correction is not performed, the self-interference channel channel gain (h _ij ) may be estimated according to Equation 2 using a Recursive Least Square (RLS) algorithm.

When the self-interference channel gain h _i is estimated according to Equation 2, the self-interference signal reconstructed from the estimated self-interference channel gain h _i is obtained as X [n] h _i [n]. In addition, the self-interference cancellation signal (y _i.DC [n]) obtained by removing the reconstructed self-interference signal (X [n] h _i [n]) from the received symbol (y _i.res [n]) is as shown in Equation 3. Is calculated.

Where N _ob is the length of (y _i.DC [n]) and is an observed symbol.

The self-interference cancellation signal (y _i.DC [n]) obtained according to Equation 3 is applied to the decoding and demodulator 250 of FIG. 2.

When the self-interference channel gain (h _i ) of Equation 2 is estimated using the RLS algorithm and the reconstructed self-interference signal (X [n] h _i [n]) is calculated, the reconstructed self-interference signal (X [n) ] h _i [n]) can reflect a nonlinear component of the power amplifier 150 of the transmitter 100 and obtain a sufficient amount of self-interference cancellation. However, as the number of antennas increases, the computational complexity increases significantly, and there is a limitation that it is difficult to apply to a communication system in reality.

However, the power amplifier 150 does not exhibit non-linear characteristics in all sections. In general, the power amplifier 150 exhibits a linear characteristic in a low power section in which the power level of the transmission signal (x _i.PA [n]) is less than or equal to a predetermined reference level, whereas a high level section in which the power level exceeds the reference level Shows nonlinear characteristics.

Therefore, since the self-interference signal component included in the received symbol y _i.res [n] is also linear in the low power period in which the power amplifier 150 exhibits linear characteristics, the self-interference channel can be easily performed according to Equation 2 with the RLS algorithm. The gain h _i may be estimated, and the estimated self-interference channel gain h _i may be applied to Equation 3 to obtain a self-interference cancellation signal y _i.DC [n].

However, in the high-level section in which the power amplifier 150 exhibits non-linear characteristics, since the self-interfering signal component included in the received symbol (y _i.res [n]) exhibits non-linear characteristics, computational complexity is increased and the RLS algorithm is applied. it's difficult.

Accordingly, in the present embodiment, a plurality of transmitters 100 are added with a pre-compensation unit 120 that performs pre-compensation using a pre-compensation function (P (·)), and the compensation function acquisition unit 240 is a pre-compensation function (P (·)) is acquired and transmitted to the preceding compensation unit 120. Therefore, the self-interference signal component included in the received symbol (y _i.res [n]) has a linear characteristic, and the compensation function obtaining unit 240 reconstructs the self-interference signal from the received symbol (y _i.res [n]). By removing (X [n] h _i [n]) easily, a self-interference cancellation signal (y _i.DC [n]) can be obtained.

However, when the compensation function obtaining unit 240 estimates the preceding compensation coefficient C _i , the preceding compensation unit 120 may already perform compensation, and the received symbol (y _i.res [n]) is compensated. It can be seen as a received symbol.

Meanwhile, the preceding compensation function P _i (·) is represented by Equation 4 as the inverse function of Equation 1.

Here, B ^-1 is an inverse function of Equation 1, c _ij ^(k) denotes the pre-compensation coefficient for the k-th nonlinear channel between the i-th transmitter and the j-th receiver, and K _PC is the preceding compensation function (P (·) ) Represents the order of pre-calibration function, and w represents the number of taps of pre-calibration function (P (·)).

From the equations 2 to 4, even if the power amplifier 150 operates in a high level section and exhibits a nonlinear characteristic, if the preceding compensation coefficient c _i and the self-interference channel gain h _i can be estimated, the preceding compensation function ( P (·)) can be obtained, and it can be seen that the self-interference signal can be easily removed by performing the compensation by the preceding compensation unit 120 using the obtained preceding compensation function (P (·)). have.

FIG. 3 shows the detailed configuration of the compensation function obtaining unit of FIG. 2, and FIG. 4 shows the detailed configuration of the preceding compensation coefficient estimating unit of FIG. 3.

Referring to FIG. 3, the compensation function obtaining unit 240 includes a preceding compensation coefficient estimating unit 241 and a preceding compensation function obtaining unit 243.

The preceding compensation coefficient estimator 241 receives the received symbol (y _i.res [n]) and the transmitted symbol (X _i [n]) to obtain the preceding compensation coefficient (c _i ) and the self-interference channel gain (h _i ). Estimate. Then, the estimated preceding compensation coefficient c _i is transmitted to the preceding compensation function acquisition unit 243.

At this time, the preceding compensation coefficient estimator 241 divides the low power operation section and the high level operation section of the power amplifier 150 as described above, thereby obtaining the preceding compensation coefficient c _i and the self-interference channel gain in each section. (h _i ) can be estimated.

Accordingly, the pre-compensation coefficient estimator 241 includes a first mode (mode1) for estimating the pre-compensation coefficient (c _i ) and the self-interference channel gain (h _i ) in a low power period in which the power amplifier 150 operates linearly. The power amplifier 150 may operate in a second mode (mode2) for estimating a preceding compensation coefficient (c _i ) and a self-interference channel gain (h _i ) in a high level period in which the power amplifier 150 operates in a nonlinear manner.

4, the preceding compensation coefficient estimator 241 includes a compensation coefficient estimator 410, a subtractor 430, a mode selection switch 450, and a first self-interference signal reconstruction unit 460 and a second self-interference. It includes a signal reconstruction unit 470.

In the first mode (mode1), the mode switch 450 selects the first mode. Accordingly, the subtractor 430 receives the received symbol (y _i.res [n]) and the first reconstructed self-interference signal (X [n] h _i ¹ [n]) obtained from the first self-interference signal reconstruction unit 460. Is accredited. And the subtractor 430 calculates the difference between the received symbol (y _i.res [n]) and the first reconstructed self-interference signal (X [n] h _i ¹ [n]) to generate an error signal (e _i [n]). ).

Here, the error signal (e _i [n]) is a self-interference cancellation signal (y _i.DC [n]) in the first mode according to Equation 3, and is _viewed as a self-interference cancellation signal (y _i.DC [n]). You can.

That is, in the first mode (mode) in which the power amplifier 150 operates linearly, since the pre-compensation unit 120 does not perform the compensation operation, the subtractor 430 of the pre-compensation coefficient estimator 241 is shown in Equation (3). Accordingly, the self-interference cancellation signal (y _i.DC [n]) is obtained and output.

Meanwhile, the first self-interference signal reconstruction unit 460 uses Equation 2 for the error signal e _i [n] output from the subtractor 430 and the transmission symbol X _i [n] applied from the transmitter 100. Using, the first self-interference channel gain (h _i ¹ ) is estimated. Then, the first reconstructed self-interference signal (X [n] h _i ¹ [n]) is obtained using the estimated first self-interference channel gain (h _i ¹ ).

In the first mode (mode1), the subtractor 430 and the first self-interference signal reconstruction unit 460 use the RLS algorithm during a predetermined number of transmission symbol (X _i [n]) intervals, to obtain the first self-interference channel gain. (h ¹ _i) it can be repeated to estimate, estimate the correct first magnetic interference channel gain (h _i ^1). That is, it is possible to improve the estimation accuracy of the first self-interference channel gain (h _i ¹ ) through iterative estimation using the RLS algorithm.

Although the subtractor 430 and the first self-interference signal reconstruction unit 460 estimate the first self-interference channel gain (h _i ¹ ) using the RLS algorithm in the first mode (mode1), the power amplifier 150 Since the section maintains the linear characteristic, the first self-interference channel gain (h _i ¹ ) can be easily obtained with low computational complexity.

Meanwhile, in the second mode, the mode selection switch 450 selects the second mode (mode2). That is, the power amplifier 150 operates with a non-linear characteristic, and thus, the preceding compensation unit 120 performs the preceding compensation.

In the second mode (mode2), the second self-interference signal reconstruction unit 470 obtains the first self-interference channel gain (h _i ¹ ) and the transmission symbol (X _i [n) obtained from the first self-interference signal reconstruction unit 460. ]) and the error signal (e _i [n]), the application receiving the second self interference channel gain (h _i ²⁾ to estimate, and the second reconstructed using the estimated second magnetic interference channel gain (h _i ²⁾ magnetic The interference signal (X [n] h _i ² [n] = ㎑ _i [n]) is obtained and output to the subtractor 430.

The second self-interference signal reconstruction unit 470 may estimate the second self-interference channel gain (h _i ² ) according to Equation 2, similarly to the first self-interference channel.

At this time, the compensation coefficient estimator 410 considers the effect of the preceding compensation unit 120 performing the preceding compensation on the received symbol (y _i.res [n]), so that the second reconstructed self-interference signal (X [n] h) _i ² [n] = ㎑ _i [n]) along with Equation 5 to estimate the preceding compensation coefficient c _i [n].

In the second mode, the compensation coefficient c _i [n] and the second self-interference channel gain h _i ² must be estimated, but these two parameters are correlated. That is, when the compensation coefficient c _i [n] is variable, the second self-interference channel gain h _i ² is also changed as the transmission signal x _i.PA [n] is changed. Conversely, when the second self-interference channel gain h _i ² is variable, the compensation coefficient c _i [n] is also changed.

Therefore, it is estimated that the compensation coefficient c _i [n] and the second self-interference channel gain h _i ² are each updated repeatedly. As an example, if the compensation coefficient c _i [n] is updated and estimated, the second self-interference channel gain h _i ² is again estimated based on the estimated compensation coefficient c _i [n]. On the other hand, if the second self-interference channel gain (h _i ² ) is updated and estimated, the compensation coefficient c _i [n] is estimated again based on the estimated second self-interference channel gain (h _i ² ).

Compensation coefficient estimation unit 410 when the compensation coefficient (c _i [n]) is estimated, and transmits the estimated compensation coefficients (c _i [n]) to obtain the prior compensation function unit (243).

Then, the preceding compensation function acquiring unit 243 obtains the preceding compensation function P (·) using the estimated compensation coefficient c _i [n] according to Equation 4, and the obtained preceding compensation function P ( ·)) To the preceding compensation unit 120 of the transmitter 100.

However, the second self-interference signal reconstruction unit 470 estimates the second self-interference channel gain (h _i ² ) in a section in which the power amplifier 150 exhibits nonlinear characteristics. Therefore, if the RLS algorithm is used as it is, computational complexity can be greatly increased.

Accordingly, in this embodiment, in order to reduce the computational complexity of the second self-interference signal reconstruction unit 470, when estimating the second self-interference channel gain (h _i ² ), forgetting factor (λ, 0 ≤λ ≤ 1 ) Is introduced, so that the error signal e _i [n], that is, the self-interference cancellation signal y _i.DC [n], is reflected at a ratio corresponding to the forgetting factor λ.

When the forgetting factor λ is introduced, the cost function C [n] of the computational complexity for the self-interference cancellation signal y _i.DC [n] may be calculated as in Equation 6.

The closer the forgetting factor (λ) is to 1, the more accurate it is because the earlier estimates are reflected and the slower the convergence rate. Therefore, the optimal forgetting factor (λ) that can quickly reach the target amount of self-interference must be searched. In addition, a high full-duplex communication capacity can be provided by setting an oblivion factor (λ) suitable for a change in an actual time-varying channel.

Accordingly, the second self-interference signal reconstruction unit 470 in the second interference channel gain estimation unit 471 so that the forgetting factor λ may be considered when estimating the second self-interference channel gain h _i ² . And an forgetting factor optimization unit 472, a step size determination unit 473, and a signal level detection unit 474. Here, the forgetting factor optimizing unit 472, the step size determining unit 473, and the signal level detecting unit 474 may be regarded as the forgetting factor setting unit.

In this embodiment, the forgetting factor setting unit receives an error signal (e _i [n]), which is a self-interference cancellation signal (y _i.DC [n]), calculates a mean square error (MSE) value, and calculates both accuracy and convergence speed. It is possible to obtain the considered forgetting factor (λ). In particular, the second self-interference channel gain (h _i ² ) can be quickly and accurately estimated by adaptively varying the forgetting factor λ according to the signal level of the required signal s _i [n].

That is, the forgetting factor λ is optimized based on the point where the signal level of the self-interference signal becomes lower than the signal level of the required signal s _i [n]. The MSE-based forgetting factor (λ) may be obtained according to Equation (7).

Here, A (·) is an MSE-based forgetting factor (λ) update function, and μ represents a step size.

5 is a view for explaining the concept of setting a standard for optimizing the forgetting factor.

5 illustrates a method of selecting an optimal forgetting factor (λ) when the signal level of the required signal s _i [n] is -70 dBm, and the X-axis direction is a self-interference cancellation signal (y _i.DC). (n)), and the Y-axis direction represents the signal level (α [n]) of the residual self-interference cancellation signal.

In order to optimize the forgetting factor λ, first, a symbol ([n ^* ]) corresponding to a signal level of a required signal s _i [n] among symbols ([n]) in the X-axis direction is searched. This can be determined by comparing the signal level α [n] of the residual self-interference cancellation signal corresponding to each symbol [n] with the signal level of the required signal s _i [n].

The signal level α [n] of the residual self-interference cancellation signal is calculated by Equation (8).

Then, a symbol ([n ^* ]) can be searched from Equation 8 to Equation 9.

When the symbol ([n ^* ]) corresponding to the signal level of the required signal s _i [n] is determined according to Equation 9, the corresponding symbol (according to the item (n> n ^* -γ) of Equation 7 ( The minimum value of the number of symbols (γ) for reflecting the forgetting factor such that the influence of the forgetting factor (λ) appears in [n ^* ]) is obtained as in FIG. 5.

As shown in Fig. 5, based on the signal level of the required signal s _i [n], optimizing the forgetting factor λ, the error signal e _i [n] can be quickly converged, yet accurate. 2 Let the self-interference channel gain (h _i ² ) be estimated.

At this time, the initial value of the forgetting factor λ may be heuristically selected according to the signal level of the required signal s _i [n].

Referring back to FIG. 4, the signal level detection unit 474 of the forgetting factor setting unit receives the error signal e _i [n], which is the self-interference cancellation signal y _i.DC [n], and desire signal (s _i [ The level of n]) is detected as in Equation 8. Then, the signal level detection unit 474 _{searches for} the signal level α [n ^* ] of the self-interference cancellation signal y _i.DC [n] satisfying Equation (11). When the symbol ([n ^* ]) of the signal level (α [n ^* ]) of the self-interference cancellation signal (y _i.DC [n]) that satisfies Equation (11) is searched, the number of symbols reflected by the forgetting factor (γ) is determined. Discriminate. Then, the step size determination unit 473 determines the step size (μ) according to the number of symbols (γ) reflected by the forgetting factor, and transmits it to the forgetting factor optimization unit 472.

The forgetting factor optimization unit 472 optimizes the forgetting factor λ according to Equation 7 and transmits the optimized forgetting factor λ to the second interference channel gain estimation unit 471.

The second interference channel gain estimator 471 acquires the second self-interference channel gain (h _i ² ) using the RLS algorithm according to Equation 2, but the forgetting factor (λ) applied by the forgetting factor optimizer 472 ) To reflect the previously estimated second self-interference channel gain (h _i ² ) by reflecting it at a predetermined rate and obtain it by repeat estimation, and reconfigure 2 self-interferences according to the obtained second self-interference channel gain (h _i ² ) The signal (X [n] h _i ² [n] = ㎑ _i [n]) is transmitted to the subtractor 430.

6 shows a self-interference cancellation method of a full-duplex communication terminal according to an embodiment of the present invention.

Referring to FIGS. 2 to 5, when the self-interference cancellation method of FIG. 6 is described, first, the terminal selects the first mode (mode1) to transmit a transmission signal (x _i.PA [n]), and transmits the received signal. Down-converted and digital-converted to obtain a received symbol (y _i.res [n]) (S10). In the first mode (mode1), the transmitter 100 outputs a transmission signal (x _i.PA [n]) so that the power amplifier 150 operates in a low level section having a linear characteristic. At this time, the preceding compensation unit 120 does not perform a preceding compensation operation for linearizing the self-interference signal.

Accordingly, the compensation function obtaining unit 240 of the receiver 200 estimates the first self-interference channel gain (h _i ¹ ) according to Equation 2 (S20). As shown in Equation 2, the compensation function acquisition unit 240 uses the estimated first self-interference channel gain (h _i ¹ ), and the first reconstructed self-interference signal (X [n] h _i ¹ [n]) ), And iterative calculation using the RLS algorithm to minimize the difference between the calculated first reconstructed self-interference signal (X [n] h _i ¹ [n]) and the received symbol (y _i.res [n]). By doing so, the first self-interference channel gain h _i ¹ can be estimated. Here is output to the first reconstructed magnetic interference signal ^{_{(X [n] h i 1}} [n]) and the received symbol (y _i.res [n]) of magnetic interference cancellation signal (y _i.DC [n]) The difference between the .

In the first mode (mode1), since the self-interference signal has a linear characteristic, the compensation function acquisition unit 240 is small in estimating the first self-interference channel gain (h _i ¹ ) using an RLS algorithm that requires iterative calculation. The first self-interference channel gain (h _i ¹ ) can be quickly estimated by calculation.

When the first self-interference channel gain (h _i ¹ ) is estimated, the terminal selects the second mode (mode2) to transmit the transmission signal (x _i.PA [n]), and down-converts and digitally converts the received signal. A reception symbol (y _i.res [n]) is obtained (S30). In the second mode (mode2), the transmitter 100 outputs a transmission signal (x _i.PA [n]) so that the power amplifier 150 operates in a high level section having a nonlinear characteristic, wherein the preceding compensator 120 Performs a preceding compensation operation for linearizing the self-interference signal using the compensation function P (·) applied from the compensation function acquisition unit 240.

The compensation function acquisition unit 240 of the receiver 200 estimates the second self-interference channel gain (h _i ² ) according to Equation 2 and compensates based on the estimated second self-interference channel gain (h _i ² ) The coefficient c _i [n] is estimated according to Equation 5 (S40). Here, the second self-interference channel gain (h _i ² ) may be estimated based on the previously estimated compensation coefficient (c _i [n]), and the initial value of the second self-interference channel gain (h _i ² ) is the first In the mode (mode1), the first self-interference channel gain (h _i ¹ ) finally estimated is set.

When the second self-interference channel gain h _i ² and the compensation coefficient c _i [n] are estimated, the compensation function acquisition unit 240 reconstructs the self-interference signal to reconstruct the second reconstructed self-interference signal X [ n] h _i ² [n] = ㎑ _i [n]), and a self-interference cancellation signal (y _i.DC [n]) is obtained from a difference from the received symbol (y _i.res [n]). (S50).

In addition, the compensation function acquiring unit 240 obtains the signal level α [n] of the residual interference cancellation signal of the self-interference cancellation signal y _i.DC [n] according to Equation 8, thereby removing the residual interference. It is determined whether the signal level α [n] of the signal converges below the signal level of the required signal s _i [n] (S60).

If the signal level α [n] of the residual interference cancellation signal converges, the compensation function acquiring unit 240 uses the estimated compensation coefficient c _i [n] to advance compensation as shown in Equation (4). The function P _i (·) is obtained, and the obtained preceding compensation function P _i (·) is transmitted to the preceding compensation unit 120 so that the nonlinearity of the self-interference signal is compensated in advance.

However, if the signal level α [n] of the residual interference signal is not converged, the second self-interference channel gain h _i ² and the compensation coefficient c _i [n] are estimated (S40). At this time, the compensation function acquiring unit 240 optimizes the forgetting factor (λ) so that the second self-interference channel gain (h _i ² ) and the compensation coefficient (c _i [n]) can be performed quickly and accurately and the optimized forgetting factor (λ). By adjusting the subsequent reflection ratio of the self-interference cancellation signal (y _i.DC [n]), the second self-interference channel gain (h _i ² ) and the compensation coefficient (c _i [n]) can be estimated quickly and accurately. To make.

Here, the forgetting factor (λ) may be optimized based on Mean Square Error (MSE). The forgetting factor (λ) searches for a symbol ([n ^* ]) corresponding to the signal level of the signal level ([alpha] [n]) of the residual interference signal and the required signal (s _i [n]), and the corresponding symbol ( It can be optimized by selecting the forgetting factor (λ) as shown in Equation (7) by using the minimum value of the number of symbols (γ) that reflects the forgetting factor that causes the effect of the forgetting factor (λ) to appear in [n ^* ]).

7 shows a simulation result of self-interference cancellation performance of a full-duplex communication terminal according to an embodiment of the present invention.

7 shows the results performed according to the simulation conditions in Table 1 in the LTE communication environment.

As illustrated in FIG. 7, the self-interference cancellation method of the full-duplex communication terminal according to the present embodiment can quickly remove the self-interference signal adaptively to the signal level of the required signal s _i [n].

The method according to the present invention can be implemented as a computer program stored in a medium for execution on a computer. Computer readable media herein can be any available media that can be accessed by a computer, and can also include any computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, and ROM (readable) Dedicated memory), RAM (random access memory), CD (compact disk) -ROM, DVD (digital video disk) -ROM, magnetic tape, floppy disk, optical data storage, and the like.

Although the present invention has been described with reference to the embodiments shown in the drawings, these are merely exemplary, and those skilled in the art will understand that various modifications and other equivalent embodiments are possible therefrom.

Therefore, the true technical protection scope of the present invention should be defined by the technical spirit of the appended claims.

BS1, BS2: Base station UE: User terminal
100: transmitter 110: coding and modulation unit
120: preceding compensation unit 130: DAC
140: up-modulation unit 150: power amplifier
200: receiver 210: low noise amplifier
220: downconverter 230: ADC
240: compensation function acquisition unit 250: decoding and demodulation unit
310: transmit antenna 310: receive antenna
241: pre-compensation coefficient estimator 240: pre-compensation function acquisition unit
410: compensation coefficient estimator 430: subtractor
450: mode selection switch 460: first self-interference signal reconstruction unit
470: second self-interference signal reconstruction unit

Claims (14)

Multiple antennas;
According to a section in which the power amplifier operates with linear and nonlinear characteristics, the amplified transmission signal is output to the plurality of antennas by dividing into a first mode and a second mode, and the nonlinear characteristics of the power amplifier are compensated in advance in the second mode. A plurality of transmitters for outputting the transmission signal by precompensating at a digital signal level according to a function; And
A reception symbol that digitally converts a received signal received through the plurality of antennas and a transmission symbol that is digitally modulated transmission data are received from the transmitter,
Based on the linearity of the self-interference signal in the first mode, the first self-interference channel gain is estimated using the received symbol and the transmission symbol, and a first reconstructed self-interference signal is obtained,
In the second mode, the second self-interference channel gain and the preceding compensation coefficient are alternately repeatedly estimated from the received symbol and the transmission symbol and the first self-interference channel gain, which are pre-compensated and transmitted, and according to the preceding compensation coefficient, A plurality of receivers acquiring a preceding compensation function and transmitting it to the transmitter; Full-duplex communication terminal comprising a. The method of claim 1, wherein the receiver
A down-conversion unit that down-converts the received signal into a baseband signal;
An ADC which converts the down-converted received signal into a digital signal and outputs the received symbol;
In the first mode, the first self-interference channel gain is repeatedly estimated according to an RLS algorithm to minimize the difference between the received symbol and the first reconstructed self-interference signal,
In the second mode, the second self-interference channel gain is estimated so that the difference between the pre-compensated received symbol and the second reconstructed self-interference signal according to the preceding compensation function is minimized, and the second obtained from the estimated second self-interference channel gain. Estimate the preceding compensation coefficient according to a known relationship between a reconstructed self-interference signal and the received symbol and the preceding compensation coefficient,
A compensation function acquiring unit acquiring an updated preceding compensation function according to the estimated preceding compensation coefficient, and transmitting the updated compensation function to the transmitter; Full-duplex communication terminal comprising a. The method of claim 2, wherein the compensation function acquisition unit
In the first mode, the first self-interference channel gain (h _i [n]) is received by receiving the received symbol (y _i.res [n]) and the transmission symbol (X [n]) from the transmitter.
Equation

(Where h _ij and M _ij are the channel gain and length of the self-interference channel (SI) between the i th transmit antenna and the j th receive antenna, respectively, φ _k (x _j ) is the basic function of the system model with self-interference. _k (x _j ) = | x _j [n] | ^k-1 x _j [n] (where x _j [n] is the transmission data of the jth transmitter), N _t is the number of antennas)
Estimate according to,
The first self-interference cancellation signal (y _i.DC [n]) that is the difference between the received symbol and the first reconstructed self-interference signal (X [n] h _i [n])
Equation

Full duplex communication terminal obtained according to. The method of claim 2, wherein the compensation function acquisition unit
In the second mode, receiving the received symbol (y _i.res [n]) and the transmitting symbol (X [n]) from the transmitter, the second self-interference channel gain (h _i [n]) is obtained.
Equation

Estimate according to,
The preceding compensation coefficient (c _i [n])
Equation

Full duplex communication terminal estimated according to. The method of claim 4, wherein the compensation function acquisition unit
Compensation function (·)
Equation

(c _ij ^(k) represents the pre-compensation coefficient for the k-th nonlinear channel between the i-th transmitter and the j-th receiver, and K _PC is the order of the pre-calibration function (P (·)) And w denotes the number of taps of pre-calibration function (P (·)).)
Full duplex communication terminal obtained according to. The method of claim 2, wherein the compensation function acquisition unit
In the second mode, an oblivion factor adjusting a reflection ratio of the self-interference cancellation signal (y _i.DC [n] or error signal (e _i [n])), which is a difference between the received symbol and the second reconstructed self-interference signal, The preceding compensation coefficient is estimated based on the reflected RLS algorithm, and the forgetting factor is a full-duplex communication terminal that is adjusted based on a Mean Square Error (MSE) method according to a signal level of desire signal (s _i [n]). The method of claim 6, wherein the compensation function acquisition unit
The signal level (α [n]) of the residual self-interference cancellation signal corresponding to the difference between the signal level of the self-interference cancellation signal and the desire signal (s _i [n]) is equal to that of the desire signal (s _i [n]). The symbol ([n ^* ]) corresponding to the signal level is searched for, and the number of symbols reflected by the forgetting factor (γ) that causes the searched symbol ([n ^* ]) to show the effect of the forgetting factor (λ) is confirmed. Factor (λ)

Full-duplex communication terminal to adjust according to. In a method for removing self-interference of a full-duplex communication terminal including a plurality of antennas, a plurality of transmitters and a plurality of receivers,
Transmitting a transmission signal in a section in which the power amplifier of the transmitter operates in a linear characteristic in the first mode, and obtaining a transmission symbol in which the transmission data is digitally modulated and a reception symbol in which the reception signal is digitally converted;
Estimating a first self-interference channel gain and obtaining a first reconstructed self-interference signal based on the linearity of the self-interference signal;
In the second mode, the power amplifier is pre-compensated for the non-linear characteristic at a digital signal level according to a pre-compensation function in a period in which the power amplifier operates with a non-linear characteristic, and transmits a transmission signal. Obtaining a digitally converted reception symbol;
Estimating a second self-interference channel gain and a pre-compensation coefficient by alternately iterating from the pre-compensated transmitted symbol and the transmission symbol and the first self-interference channel gain; And
Obtaining the preceding compensation function according to the preceding compensation coefficient and transmitting it to the transmitter; Self-interference cancellation method of a full-duplex communication terminal comprising a. 10. The method of claim 8, wherein obtaining the first reconstructed self-interference signal is
A self-interference cancellation method of a full-duplex communication terminal to obtain a first reconstructed self-interference signal by repeatedly estimating a first self-interference channel gain according to an RLS algorithm so that the difference between the received symbol and the first reconstructed self-interference signal is minimized. 10. The method of claim 9, wherein obtaining the first reconstructed self-interference signal is
The first self-interference channel gain (h _i [n]) is received by receiving the received symbol (y _i.res [n]) and the transmitting symbol (X [n]) from the transmitter.
Equation

(Where h _ij and M _ij are the channel gain and length of the self-interference channel (SI) between the i th transmit antenna and the j th receive antenna, respectively, φ _k (x _j ) is the basic function of the system model with self-interference. _k (x _j ) = | x _j [n] | ^k-1 x _j [n] (where x _j [n] is the transmission data of the jth transmitter), N _t is the number of antennas)
Estimate according to,
The first self-interference cancellation signal (y _i.DC [n]) that is the difference between the received symbol and the first reconstructed self-interference signal (X [n] h _i [n])
Equation

A self-interference cancellation method of a full-duplex communication terminal acquired according to. The method of claim 10, wherein estimating the preceding compensation coefficient is
Estimating a second self-interference channel gain such that a difference between a previously-compensated received symbol and a second reconstructed self-interference signal is minimized according to the preceding compensation function; And
Estimating the preceding compensation coefficient according to a predetermined relationship between the second reconstructed self-interference signal obtained from the estimated second self-interference channel gain and the received symbol and the preceding compensation coefficient; Self-interference cancellation method of a full-duplex communication terminal comprising a. The method of claim 11, wherein estimating the preceding compensation coefficient is
Equation

Estimate according to,
The preceding compensation coefficient (c _i [n])
Equation

A self-interference cancellation method of a full-duplex communication terminal estimated according to. The method of claim 12, wherein the step of transmitting to the transmitter
Compensation function (·)
Equation

(c _ij ^(k) represents the pre-compensation coefficient for the k-th nonlinear channel between the i-th transmitter and the j-th receiver, and K _PC is the order of the pre-calibration function (P (·)) And w denotes the number of taps of pre-calibration function (P (·)).)
A self-interference cancellation method of a full-duplex communication terminal obtained according to the method. The method of claim 11, wherein estimating the preceding compensation coefficient is
The self-interference cancellation signal (y _i.DC [n] or error signal (e _i [n])) which is the difference between the received symbol and the second reconstructed self-interference signal when estimating the preceding compensation coefficient based on the RLS algorithm in the second mode. ) Is estimated by applying the forgetting factor to adjust the reflection ratio, and the forgetting factor is based on the Mean Square Error (MSE) method according to the signal level of the desire signal (s _i [n]), and is adjusted. How to remove self-interference.

Images