KR102029738B1 - Channel estimation apparatus and method - Google Patents

Abstract

The channel estimating apparatus includes a pilot symbol extracting unit, a least squares channel estimating unit, a Kalman filter, and a Wiener filter. The pilot symbol extractor extracts a plurality of pilot symbols from the symbols of the plurality of frequency domains. The least square channel estimator compares the plurality of extracted pilot symbols with the plurality of reference pilot symbols to generate a plurality of least square channel frequency responses. Kalman filter Kalman-filters a plurality of least squares channel frequency responses. The winner filter filters the plurality of Kalman-filtered least squares channel frequency responses.

Description

Channel estimation apparatus and method {CHANNEL ESTIMATION APPARATUS AND METHOD}

The present invention relates to an orthogonal frequency division multiplexing (OFDM) based signal receiving apparatus and method. In particular, the present invention relates to a channel estimating apparatus in a signal receiving apparatus and a method therefor.

As the paradigm of wireless communication service shifts from voice service center to multimedia service center, Orthogonal Frequency Division Multiplexing (OFDM) technology and Orthogonal Frequency Division Multiple Access (OFDMA) technology using it are noted. I got it. This OFDM method has high bandwidth-efficiency and strong characteristics for multipath fading, so it is a 4th generation mobile communication system, terrestrial digital television, and wireless local area network. , WLAN) and the like.

Next, a conventional OFDM signal transmission device and a conventional OFDM signal reception device will be described with reference to FIG. 1.

1 is a block diagram illustrating a conventional OFDM signal transmission device and an OFDM signal reception device.

As shown in FIG. 1, the conventional OFDM signal transmission apparatus 10 includes a digital modulator 11, an Inverse Fast Fourier Transform calculator (IFFT calculator) 13, and a Cyclic Prefix Inserter ( CP Insertion Unit 15, Digital-to-Analog Converter (DAC) 16, Transmitter 17, and Transmit Antenna 18 are included.

The digital modulator 11 performs digital modulation such as binary phase shift keying (BPSK), quadrature amplitude modulation (QAM), 16-QAM, 64-QAM, and the like, on data symbols that are binary data for transmission on a channel. Create a digital-modulated symbol. The digital modulator 11 may also be referred to as a symbol mapping unit.

The IFFT calculator 13 performs IFFT on the plurality of digital modulation symbols output from the digital modulator 11 to generate a plurality of inverse-fast-fourier-transformed symbols.

The CP inserter 15 adds a signal called Cyclic Prefix to the front of the plurality of IFFT symbols output from the IFFT operator 13 to generate a symbol group into which the CP is inserted. Here, the Cyclic Prefix means symbols of the rear part of the plurality of IFFT symbols.

The DAC 16 receives a symbol group into which the CP is inserted from the CP inserting unit 15 and converts the symbol group into analog to generate one OFDM symbol.

The transmitter 17 amplifies and converts an OFDM symbol generated by the DAC 16 into a radio frequency (RF) signal and transmits the converted OFDM symbol to a channel through the transmit antenna 18.

Meanwhile, as shown in FIG. 1, the conventional OFDM signal receiving apparatus 20 includes a receiving antenna 21, a receiving unit 22, and an analog-to-digital converter (ADC) 23. ), A Cyclic Prefix remover 24 (hereinafter referred to as CP remover) 24, a Fast Fourier Transform calculator (FFT calculator) 26, and a digital demodulator 28.

The receiver 22 receives an OFDM symbol from the channel through the reception antenna 21.

The ADC 23 converts the OFDM symbols received by the receiver 22 into digital to generate a plurality of digital symbols.

The CP remover 24 removes the Cyclic Prefix from the plurality of digital symbols generated by the ADC 23.

The FFT calculator 26 receives a plurality of symbols output from the CP remover 24 and performs fast Fourier transform to generate a plurality of frequency domain symbols.

The digital demodulator 28 digitally demodulates a plurality of frequency domain symbols output from the FFT calculator 26 to generate binary data symbols.

Meanwhile, the OFDM signal transmitted by the signal transmission device arrives at the signal reception device through a multipath. Therefore, the OFDM signal is distorted, and the signal receiving apparatus needs to measure and compensate for the degree of distortion. Measuring the distortion of such a signal is called channel estimation, and a complicated channel estimation process is required for accurate channel estimation.

SUMMARY OF THE INVENTION The present invention provides a channel estimating apparatus, a channel estimating method, a signal receiving device, and a signal receiving method for effectively demodulating a signal through accurate channel estimation while reducing complexity.

In accordance with an aspect of the present invention, a channel estimating apparatus includes: a Fourier transform unit generating symbols on a plurality of frequency axes; A Kalman filter for generating a Kalman-filtered plurality of channel frequency responses by Kalman-filtering a plurality of least squares channel frequency responses; A discrete inverse Fourier transform unit for generating a plurality of channel impulse responses by discrete inverse Fourier transforming the Kalman-filtered plurality of channel frequency responses; A discrete Fourier transform unit for generating a plurality of discrete Fourier transformed channel frequency responses by performing discrete Fourier transform on the plurality of channel impulse responses and a predetermined number of null symbols; A winner filter for winner-filtering the plurality of discrete Fourier transformed channel frequency responses; And a channel compensator for channel-compensating symbols on the plurality of frequency axes based on the winner-filtered plurality of channel frequency responses. The Kalman filter calculates an estimated value of the channel frequency response at time n based on the measured value of the channel frequency response at time n-1, and uses the error between the calculated estimate and the measured value of time n. Compute the correction value by correcting the estimated value of. The channel compensator channel-compensates the symbols on the plurality of frequency axes using the correction value.

A channel estimating apparatus; a pilot symbol extracting unit extracting a plurality of pilot symbols from symbols of a plurality of frequency domains; And a least square channel estimator configured to generate the plurality of least square channel frequency responses by comparing the extracted plurality of pilot symbols with a plurality of reference pilot symbols.

A channel estimation method according to another embodiment of the present invention includes the steps of generating symbols on a plurality of frequency axes; Kalman-filtering a plurality of least squares channel frequency responses to produce the Kalman-filtered plurality of channel frequency responses; Generating a plurality of channel impulse responses by discrete inverse Fourier transforming the Kalman-filtered plurality of channel frequency responses; Generating a discrete Fourier transformed channel frequency response by performing discrete Fourier transform on the plurality of channel impulse responses and a predetermined number of null symbols; Winner-filtering the discrete Fourier transformed plurality of channel frequency responses; And channel-compensating the symbols on the plurality of frequency axes based on the winner-filtered plurality of channel frequency responses. The generating of the Kalman-filtered plurality of channel frequency responses comprises: calculating an estimated value of the channel frequency response at time n based on the measured value of the channel frequency response at time n-1; And calculating a correction value by correcting the estimated value of the viewpoint n using the error between the calculated estimated value and the measured value of the viewpoint n. Channel-compensating the symbols on the plurality of frequency axes comprises channel-compensating the symbols on the plurality of frequency axes using the correction value.

The channel estimation method includes extracting a plurality of pilot symbols from a plurality of symbols in a frequency domain; And comparing the extracted plurality of pilot symbols with a plurality of reference pilot symbols to generate the plurality of least square channel frequency responses.

The signal receiving apparatus according to the embodiment of the present invention can effectively demodulate a signal through accurate channel estimation while maintaining low complexity.

1 is a block diagram illustrating a conventional OFDM signal transmission device and an OFDM signal reception device.
2 is a block diagram of an OFDM signal receiving apparatus according to an embodiment of the present invention.
3 is a block diagram of a channel estimator according to an embodiment of the present invention.
4 is a flowchart illustrating a signal receiving method according to an embodiment of the present invention.
FIG. 5 is a graph showing bit error rate (BER) performance according to various channel estimation methods in a Rayleigh EPA channel having a maximum Doppler frequency of 5 Hz.
6 shows various channels in a Rayleigh EVA channel with a maximum Doppler frequency of 70 Hz.
7 is a graph showing BER performance according to various channel estimation methods in a Rayleigh ETU channel having a maximum Doppler frequency of 300 Hz.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

In addition, when a part is said to "include" a certain component, which means that it may further include other components, except to exclude other components unless otherwise stated.

Hereinafter, an OFDM signal receiving apparatus 100 according to an embodiment of the present invention will be described with reference to the drawings.

2 is a block diagram of an OFDM signal receiving apparatus according to an embodiment of the present invention.

As shown in FIG. 2, the OFDM signal receiving apparatus 100 according to the embodiment of the present invention includes a receiving antenna 110, a receiving unit 120, an ADC 130, a CP removing unit 140, and an FFT calculating unit. 160, a channel estimator 180, a channel compensator 190, and a digital demodulator 195.

The receiver 120 receives an OFDM symbol from a channel through the reception antenna 110.

The ADC 130 converts the OFDM symbols received by the receiver 120 into digital to generate a plurality of digital symbols.

The CP remover 140 removes the cyclic prefix from the plurality of digital symbols generated by the ADC 130.

The FFT calculator 160 receives a plurality of CP-removed symbols output from the CP remover 140 and performs fast Fourier transform to generate symbols on a plurality of frequency axes.

The channel estimator 180 estimates a channel based on the plurality of CP removed symbols.

The channel compensator 190 generates a plurality of channel-compensated symbols by performing channel compensation on the plurality of CP-removed symbols output from the CP remover 140 based on the estimated channel. do.

The digital demodulator 195 generates final data symbols by performing digital demodulation on the plurality of channel-compensated symbols output from the channel compensator 180.

3 is a block diagram of a channel estimator according to an embodiment of the present invention.

As shown in FIG. 3, the channel estimator 180 according to an exemplary embodiment of the present invention includes a pilot symbol extractor 181, a least square channel estimator 182, a Kalman filter 183, and an IDFT calculator 184. , A DFT calculator 185, a Wiener filter 186, and an interpolator 187.

The receiver 120 receives an OFDM symbol from a channel through the reception antenna 110.

The ADC 130 converts the received OFDM symbol into digital to generate a plurality of digital symbols.

The CP remover 140 removes the Cyclic Prefix from the plurality of digital symbols.

The FFT calculator 160 receives a plurality of CP-removed symbols and performs fast Fourier transform to generate symbols on a plurality of frequency axes.

The pilot symbol extractor 181 extracts a plurality of pilot symbols from the symbols on the plurality of frequency axes output by the FFT calculator 160.

The least square channel estimator 182 generates a plurality of least square channel frequency responses by comparing the plurality of extracted pilot symbols with the plurality of reference pilot symbols.

Kalman filter 183 generates Kalman-filtered channel frequency responses by Kalman filtering the plurality of least squares channel frequency responses.

The IDFT operator 184 generates IDs by performing IDFT with a plurality of Kalman filtered channel frequency responses.

The DFT operator 185 generates a plurality of channel frequency responses by performing a DFT with a plurality of channel impulse responses output from the IDFT operator 184 and a predetermined number of null symbols.

The winner filter 186 generates a plurality of winner-filtered channel frequency responses H _WF by Wiener filtering the plurality of channel frequency responses output by the DFT calculator 185.

The interpolator 187 interpolates the plurality of Wiener-filtered channel frequency responses to generate a plurality of interpolated channel frequency responses for the entire subcarrier.

The channel compensator 180 generates a plurality of channel-compensated symbols by channel-compensating a plurality of CP-removed symbols output by the CP remover 140 based on the estimated channel.

The digital demodulator 190 generates binary data symbols by performing digital demodulation on the plurality of channel-compensated symbols output from the channel compensator 180.

4 is a flowchart illustrating a signal receiving method according to an embodiment of the present invention.

The receiver 120 receives an OFDM symbol from a channel through the reception antenna 110 (S101).

The ADC 130 converts the received OFDM symbol into digital to generate a plurality of digital symbols (S103).

The CP remover 140 removes the Cyclic Prefix from the plurality of digital symbols (S105).

The FFT calculator 160 receives a plurality of CP-removed symbols and performs fast Fourier transform to generate symbols on a plurality of frequency axes (S107).

The pilot symbol extracting unit 181 may include a plurality of pilot symbols (y _n (1), y _n (2), .., y _n (L) from symbols on the plurality of frequency axes output by the FFT calculating unit 160 at time n. )) Is extracted (S109). L represents the number of pilot symbols.

The least square channel estimator 182 compares the plurality of extracted pilot symbols with a plurality of reference pilot symbols (x _n (1), x _n (2), .., x _n (L)) to determine a plurality of least squares. Generate a channel frequency response (S111). The plurality of least squares channel frequency responses may be expressed by Equation 1 below.

The Kalman filter 183 generates Kalman-filtered channel frequency responses by Kalman filtering the plurality of least squares channel frequency responses (S113). Hereinafter, an example of the Kalman filter 183 will be described using an equation.

First, the Kalman filter 183 predicts the state estimation value of the channel frequency response at time n based on the measured value of the channel frequency response at time n-1. The Kalman filter 183 may deductively calculate a state estimate as shown in Equation 2.

In Equation 2, Fn represents a state transition matrix based on the state of time n-1 at time n, and Fn = aI. a may be expressed as in Equation 3.

In Equation 3, J ₀ () is a Bessel function of the first kind, and f _D T _S has a normal Doppler frequency.

은 수학식 4와 같은 공분산 행렬 Qn를 가지는 다변수 정규 분포 잡음 변수이다.In Equation 2,

Is a multivariate normal distribution noise variable having a covariance matrix Qn as shown in Equation 4.

는 채널의 전력 지연 값이다.In Equation 4,

Is the power delay value of the channel.

The Kalman filter 183 then predicts the estimated value P _{n | n-1} of the state covariance matrix at time n based on the measured value of the pilot symbol at time _n-1 . The Kalman filter 183 may deduce the estimated value P _{n | n−1} of the state covariance matrix as shown in Equation 5 below.

The Kalman filter 183 then inductively corrects the predicted value at time n using the error between the predicted value at time n and the measured value at time n.

Meanwhile, at time point n, the received symbol may be represented by a linear regression model as shown in Equation 6 below.

Accordingly, the Kalman filter 183 can obtain an error e _n at the time point n as shown in Equation 7 below.

는 시점 n에서의 파일롯 심볼의 측정값 벡터이고, Xn은 레퍼런스 파일롯 심볼의 벡터이며,

는 시점 n에서의 파일롯 심볼의 예측값 벡터이다.In Equation 7,

Is a vector of measurements of the pilot symbols at time point n, Xn is a vector of reference pilot symbols,

Is a predictive vector of pilot symbols at time point n.

Meanwhile, the Kalman filter 183 calculates an optimal Kalman gain as shown in Equation 8 below.

In equation (8), r _n is equal to equation (9).

)을 귀납적으로 보정하여 보정 값(

)을 구한다. 칼만 필터(183)는 이 보정값을 출력한다.Thereafter, the Kalman filter 183 calculates an estimated value of the channel at time point n as shown in Equation 10 below.

) Inductively compensates for correction values (

) The Kalman filter 183 outputs this correction value.

The Kalman filter 183 inductively corrects the estimated value P _{n | n−1} of the state covariance matrix at time point n as shown in Equation 11 below to obtain a correction value P _{n | n} .

The IDFT operator 184 generates IDs by performing IDFT with a plurality of Kalman filtered channel frequency responses (S115).

The DFT operator 185 generates a plurality of channel frequency responses by performing a DFT with a plurality of channel impulse responses output from the IDFT operator 184 and a predetermined number of null symbols (S117). As a result, tabs having an index larger than L may be eliminated.

The winner filter 186 generates a plurality of winner-filtered channel frequency responses H _WF by Wiener filtering the plurality of channel frequency responses output by the DFT calculator 185 (S119).

In one embodiment, the Wiener filter 186 may generate a plurality of Wiener-filtered channel frequency responses (H _WF ) by Wiener filtering in one dimension as shown in Equation (12).

In Equation 12, R _hp represents a cross-correlation matrix between all subcarriers in one OFDM symbol and subcarriers having a reference signal in the OFDM symbol, and R _hh represents an autocorrelation matrix of subcarriers having a reference signal in one OFDM symbol.

In yet another embodiment, the Wiener filter 186 may generate Wiener-filtered channel frequency responses H _WF by Wiener filtering in two dimensions, as shown in Equations 13 and 14.

In Equation 13, R _hp ^f represents a cross-correlation matrix between all subcarriers in one OFDM symbol in the frequency domain and subcarriers with reference signals in the OFDM symbol, and R _hh ^f represents a reference signal in one OFDM symbol in the frequency domain _{Denotes an} autocorrelation matrix of subcarriers with < RTI ID = 0.0 > H _WF ^f < / RTI >

In Equation 14, R _hp ^t represents a cross-correlation matrix between all subcarriers in one OFDM symbol in the time domain and subcarriers with reference signals in the OFDM symbol, and R _hh ^t is a reference signal in one OFDM symbol in the time domain. _{Denotes an} autocorrelation matrix of subcarriers with H _HF denotes a plurality of Wiener-filtered channel frequency responses.

The interpolator 187 interpolates the plurality of Wiener-filtered channel frequency responses to generate a plurality of interpolated channel frequency responses for all subcarriers (S121). In particular, the interpolator 187 may generate a plurality of interpolated channel frequency responses by interpolating on the sub-frequency axis and the time axis.

The channel compensator 180 generates a plurality of channel-compensated symbols by channel-compensating the plurality of CP-removed symbols output from the CP remover 140 based on the estimated channel ( S123).

The digital demodulator 190 generates final data symbols by performing digital demodulation on the plurality of channel-compensated symbols output from the channel compensator 180 (S125).

Next, the performance of the channel estimation apparatus according to an embodiment of the present invention will be described with reference to FIGS. 5 to 7.

FIG. 5 is a graph showing bit error rate (BER) performance according to various channel estimation methods in a Rayleigh EPA channel having a maximum Doppler frequency of 5 Hz.

In particular, the graph of FIG. 5 was measured in the parameter environment in Table 1 below.

Bandwidth 20 MHz Sample frequency 30.72 MHz Subframe duration 1 ms Subccarier spacing 15 kHz FFT size 2048 Occupied subcarriers 1200 + DC subcarrier = 1201 No. subcarriers / RB 12 No. of RB's / subframe 100 CP size (samples) 512 (extended CP) No. of OFDM symbols / subframe 12 (extended CP) No. of reference signals per RB 8 Modulation scheme QPSK Noise AWGN No. of antennas 1x1 Channel estimation techniques LS with linear interpolation, MMSE, Wiener Filter, Creative CE Channel models 3GPP LTE extended channel models: EPA, EVA, ETU

As shown in Fig. 5, in a Rayleigh EPA channel having a maximum Doppler frequency of 5 Hz, the BER performance of the embodiment of the present invention is better than the BER performance of LS, and is similar to the BER performance of MMSE and WF.

6 is a graph showing BER performance according to various channel estimation methods in a Rayleigh EVA channel having a maximum Doppler frequency of 70 Hz.

As shown in FIG. 6, in the Rayleigh EVA channel having a maximum Doppler frequency of 70 Hz, the BER performance of the embodiment of the present invention is better than that of LS and MMSE, and is similar to the BER performance of WF.

7 is a graph showing BER performance according to various channel estimation methods in a Rayleigh ETU channel having a maximum Doppler frequency of 300 Hz.

As shown in FIG. 7, in the Rayleigh ETU channel having the maximum Doppler frequency of 300 Hz, the BER performance of the embodiment of the present invention is better than that of the LS, MMSE, and WF.

Features, structures, effects, and the like described in the above embodiments are included in at least one embodiment of the present invention, and are not necessarily limited to only one embodiment. Furthermore, the features, structures, effects, and the like illustrated in the embodiments may be combined or modified with respect to other embodiments by those skilled in the art to which the embodiments belong. Therefore, contents related to such combinations and modifications should be construed as being included in the scope of the present invention.

Although the above description has been made based on the embodiments, these are merely examples and are not intended to limit the present invention. Those skilled in the art to which the present invention pertains may not have been exemplified above without departing from the essential characteristics of the present embodiments. It will be appreciated that many variations and applications are possible. For example, each component specifically shown in the embodiment can be modified. And differences relating to such modifications and applications will have to be construed as being included in the scope of the invention defined in the appended claims.

Claims (15)

Fourier transform unit for generating a plurality of symbols on the frequency axis;
A Kalman filter for generating a Kalman-filtered plurality of channel frequency responses by Kalman-filtering a plurality of least squares channel frequency responses;
A discrete inverse Fourier transform unit for generating a plurality of channel impulse responses by discrete inverse Fourier transforming the Kalman-filtered plurality of channel frequency responses;
A discrete Fourier transform unit for generating a plurality of discrete Fourier transformed channel frequency responses by performing discrete Fourier transform on the plurality of channel impulse responses and a predetermined number of null symbols;
A winner filter for winner-filtering the plurality of discrete Fourier transformed channel frequency responses; And
A channel compensator for channel-compensating symbols on the plurality of frequency axes based on the winner-filtered plurality of channel frequency responses
Including,
The Kalman filter,
Calculate an estimated value of the channel frequency response at time n based on the measurement of the channel frequency response at time n-1,
Calculates a correction value by correcting the estimated value of the time point n using the error between the calculated estimated value and the measured value of the time point n,
The channel compensator,
Channel-compensating symbols on the plurality of frequency axes using the correction value.
Channel estimation device. The method of claim 1,
A pilot symbol extracting unit for extracting a plurality of pilot symbols from a plurality of symbols of frequency domains; And
And a least square channel estimator for generating the plurality of least square channel frequency responses by comparing the extracted plurality of pilot symbols with a plurality of reference pilot symbols.
Channel estimation device. delete The method of claim 1,
And an interpolator configured to generate the interpolated plurality of channel frequency responses by interpolating the winner-filtered plurality of channel frequency responses.
Channel estimation device. The method of claim 4, wherein
Further comprising a CP removal unit for removing the Cyclic Prefix from the plurality of digital symbols,
The Fourier transform unit,
Generating a symbol on the plurality of frequency axes by performing fast Fourier transform on the plurality of CP-removed symbols
Channel estimation device. delete The method of claim 1,
And a digital demodulator for digitally demodulating the plurality of channel-compensated symbols to generate binary data symbols.
Channel estimation device. The method of claim 5,
A receiving unit for receiving an OFDM symbol from a channel through a receiving antenna;
Further comprising an analog-to-digital converter for converting the OFDM symbol into the plurality of digital symbols
Channel estimation device. Generating a symbol on a plurality of frequency axes;
Kalman-filtering a plurality of least squares channel frequency responses to produce the Kalman-filtered plurality of channel frequency responses;
Generating a plurality of channel impulse responses by discrete inverse Fourier transforming the Kalman-filtered plurality of channel frequency responses;
Generating a discrete Fourier transformed channel frequency response by performing discrete Fourier transform on the plurality of channel impulse responses and a predetermined number of null symbols;
Winner-filtering the discrete Fourier transformed plurality of channel frequency responses; And
Channel-compensating symbols on the plurality of frequency axes based on the winner-filtered plurality of channel frequency responses
Including,
Generating the Kalman-filtered plurality of channel frequency responses,
Calculating an estimated value of the channel frequency response at time n based on the measured value of the channel frequency response at time n-1; And
Calculating a correction value by correcting the estimated value of the time point n using an error between the calculated estimated value and the measured value of the time point n,
Channel-compensating the symbols on the plurality of frequency axes includes:
Channel-compensating symbols on the plurality of frequency axes using the correction value;
Channel estimation method. The method of claim 9,
Extracting a plurality of pilot symbols from the symbols of the plurality of frequency domains; And
Generating the plurality of least squares channel frequency responses by comparing the extracted plurality of pilot symbols with a plurality of reference pilot symbols.
Channel estimation method. delete The method of claim 9,
Interpolating the winner-filtered plurality of channel frequency responses to generate the interpolated plurality of channel frequency responses.
Channel estimation method. The method of claim 12,
Receiving an OFDM symbol from a channel via a receive antenna;
Converting the received OFDM symbol into a plurality of digital symbols;
Removing a cyclic prefix from the converted plurality of digital symbols;
Performing fast Fourier transform on the plurality of CP-removed symbols to generate symbols on the plurality of frequency axes;
Channel estimation method. The method of claim 13,
Channel-compensating the plurality of CP-removed symbols based on the interpolated plurality of channel frequency responses to generate the channel-compensated plurality of symbols.
Channel estimation method. The method of claim 14,
Performing digital demodulation on the plurality of channel-compensated symbols to generate binary data symbols;
Channel estimation method.

Images