KR20010062996A - Channel equalizer of a HDTV - Google Patents

Abstract

PURPOSE: A channel equalizer for an HDTV(high definition TV) is provided to increase convergence speed and improve efficiency in a multipath channel environment by equalizing the input signal size attenuation and phase distortion respectively. CONSTITUTION: A tap coefficient renewal module(40) renews the tap coefficient for equalizing the magnitude and phase of the signal. A storing module(70) stores the renewed signal magnitude and phase equalization tap coefficient and outputs them depending on the pilot signal location data. A linear interpolation module(80) interpolates the signal magnitude and phase equalization tap coefficient. A first and a second operation module(90,91) take sine and cosine functions for the phase-corrected signals. A complex processing module(120) generates a complex number by multiplying the corrected signal magnitude with the sine and cosine functions. A control module(130) provides the location data of the pilot signal to an equalizing module, the storing module(70), the tap coefficient renewal module(40) and the linear interpolation module(80).

Description

Channel equalizer of high definition television

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a channel equalizer for digital high definition television (hereinafter referred to as HDTV), and in particular, a slow convergence speed LMS (Least Mean Square) in a channel equalization technique for a European type digital terrestrial broadcasting COFDM (Coded Orthogonal Frequency Division Multiplexing) system. Unlike the technique, the present invention relates to a channel equalizer of HDTV for obtaining faster convergence and improved performance in a multipath channel environment by equalizing the magnitude and phase of a received signal independently using a pilot signal.

Modulation method that parallelizes data, orthogonal frequency division (OFDM) transmission method, which is one of multi-carrier modulation (MCM) in block units, and then transmits the data in parallel through a plurality of subcarriers within a given frequency band. to be. In this case, in order to increase frequency efficiency, the distance between subcarriers is set to 1 / T which is the minimum distance that can maintain orthogonality between subcarriers. Here, T is a symbol period between subcarriers.

One of the advantages of this OFDM system is to divide the frequency selective channel generated in the multipath channel environment into several small subchannels and approximate each subchannel as a frequency non-selective channel. It can be done. Therefore, unlike single-carrier modulation (SCM) which requires a complex channel equalizer in a frequency selective channel, the OFDM transmission scheme can be sufficiently equalized with a simple channel equalizer having one tap.

However, simply transmitting data in parallel to multiple subcarriers cannot expect a low bit error rate in a frequency selective channel. This is because data transmitted on a subcarrier having a low signal-to-noise ratio (SNR) has a high symbol error.

In order to make up for such drawbacks, a Forward Error Correction (FEC) method is generally used, and such an OFDM system is called a Coded Orthogonal Frequency Division (COFDM) system.

In general, the physical channel of a high-speed digital communication system including HDTV is free space. This channel can be considered as an ideal channel when there is no vacuum in the medium, but the actual transmission signal is received through multiple paths due to the deflection of the path propagated by weather changes such as air, clouds, atmospheric layers, or the effects of the sun. do.

In addition, a fixed multipath, which is reflected, transmitted and propagated by a terrain, a mountain, a cliff, a land, and the like, and a time-varying multipath generated by an airplane, a vehicle, or the like, may occur. Since the multipath propagation is transmitted through multiple paths having different propagation times, the inter-symbol interface, which is the biggest factor that degrades the performance of a high-speed digital communication system in transmitting digital signals, is used. Will cause: ISI).

In this way, the channel does not satisfy the ideal characteristics due to various causes, and distortion occurs in transmitting a signal.

In a digital transmission system such as HDTV, a signal distortion causes a bit detection error on the receiving side, which may result in a phenomenon that the entire screen cannot be restored or an entirely different image appears.

In order to overcome this phenomenon, a channel equalizer is called a channel equalizer that compensates for the characteristics of a channel by processing a received signal passing through a distorted channel out of an ideal characteristic, thereby reducing a bit detection error.

In order to perform channel equalization from the received signal, there is a method of estimating the distortion degree of the channel by comparing the waveform of the distorted data pulse with the original waveform through the channel at the transmitting side.

1 is a diagram illustrating a transmission signal structure of a European digital HDTV, and illustrates a structure of pilot signals and data transmitted in a standard of a European digital HDTV.

In FIG. 1, the portions marked with black dots are pilot signals, and the white dots represent data signals. The pilot signals are modulated at 4/3 + j0 and transmitted with real values. Here, the horizontal axis is a frequency component for each OFDM symbol, which satisfies 1705 or 6817 depending on the mode, and the vertical axis represents the time axis of successive OFDM symbols.

Next, the channel equalization method using the LMS algorithm, which is used a lot, will be described.

The LMS equalization technique is the most common channel equalization technique used in OFDM systems and other systems. It is a steepest equalization technique that repeatedly updates tap coefficients to minimize mean square error (MSE). As mentioned above, since each subchannel of the OFDM system can be approximated as a frequency non-selective channel, the i-th subchannel can be represented as shown in FIG. 2. 2 is a diagram illustrating a structure of a channel equalizer using a conventional LMS equalization technique, and includes a channel composed of a multiplier and an adder and an equalizer composed of a multiplier. That is, when a transmission signal is input, the channel multiplier multiplies the input transmission signal by a complex number of channel equalization response coefficients and transfers the result to the adder.

The adder of the channel adds a white Gaussian noise signal to the signal output from the multiplier and inputs it to the equalizer on the receiving side. Thus, the equalizer multiplies the signal passing through the channel by the tap coefficient of the channel equalizer, thereby outputting the equalized signal. Where, at time i, each complex signal x _i is a transmitted signal, α _i is a complex channel response coefficient, n _i is an additive white Gaussian noise (AWGN), y _i is a received signal, w _i is the tap coefficient of the channel equalizer, and z _i is the equalized signal.

When the transmitted signal x _i is a pilot signal, the tap coefficient of the channel equalizer is updated in a direction that minimizes the mean square error. Therefore, the cost function J used is given by Equation 1 below.

The LMS equalization technique takes a gradient with respect to the channel equalizer tap coefficient in Equation 1 to obtain a tap coefficient that minimizes the cost function J. That is, when the inclination is taken as in Equation 2 below.

Accordingly, the tap coefficient update expression of the LMS method is as shown in Equation 3 below.

Where μ is a positive constant representing the tap coefficient update magnitude.

The output signal of the equalizer using this updated tap coefficient is expressed by Equation 4 below.

In the LMS method according to the related art, the attenuation components and the phase distortion of the signal magnitude are equalized by one tap coefficient update equation. Therefore, there is a problem in that a slow convergence characteristic and a hardware implementation are complicated in a poor multipath environment, and there is a problem in that information on a channel size response required for a Viterbi decoder installed at the rear of the channel equalizer cannot be accurately provided.

Therefore, the present invention has been made to solve the above-mentioned problems according to the prior art, the object of the present invention, unlike the LMS method, by independently equalizing the input signal magnitude attenuation and phase distortion, respectively, a poor multi-path channel It is to provide HDTV channel equalizer for faster convergence and improved performance in the environment.

Another object of the present invention is to provide a channel equalizer that simplifies the implementation of hardware so that all subchannels can be equalized sequentially with one channel equalizer.

A feature of the present invention for achieving the above object is to configure one channel equalizer capable of independently equalizing signal magnitude equalization and phase equalization respectively.

In addition, another feature of the present invention is to update the tap coefficient of the equalizer using a pilot signal previously promised at the transmitting and receiving sides in the transmission signal structure, and to use linear interpolation for data signals between pilot signals. To equalize.

That is, the characteristic of the channel equalizer of the HDTV according to the present invention is that in the channel equalizer of the HDTV, a tap for magnitude and phase equalization of a signal using an equalized signal, a pilot signal, and a previous tap coefficient value is used. A tap coefficient updating unit for updating the coefficients; A storage unit for storing the updated signal size and the phase equalization tap coefficient, respectively, and outputting the received signal signal according to the provided pilot signal position information; A linear interpolator configured to linearly interpolate each of the signal size equalization tap coefficients and phase equalization tap coefficients stored in the storage unit according to pilot position information; First and second calculation units which respectively take a sine function and a cosine function for the phase-corrected signal in the linear interpolator; A complex number processing unit for generating a complex number by multiplying each result of the sine function and the cosine function by the first and second calculation units by the signal magnitude value corrected by the linear interpolation unit; And an equalizer for outputting an equalization value by multiplying the complex number processed by the complex number processor and an input value, and a controller for providing position information of the pilot signal to the storage unit, the tap coefficient updater, and the linear interpolator, respectively.

1 is a diagram showing a transmission signal structure of a European digital high definition television;

2 is a view showing the structure of a channel equalizer of a high definition television using the LMS equalization technique according to the prior art;

3 is a view showing the structure of a channel equalizer according to the present invention;

4 shows the role of the signal size equalizer and the phase equalizer;

5 is a block diagram of a channel equalizer of high definition television according to the present invention;

<Explanation of symbols for main parts of drawings>

30, 100, 110: multiplier 40: tap coefficient update unit

50, 60: adder 70: tap coefficient storage unit

80: linear interpolation section 90, 91: calculation section

Hereinafter, a preferred embodiment of a channel equalizer according to the present invention will be described in detail with reference to the accompanying drawings.

3 is a view showing the structure of a channel equalizer according to the present invention.

First, in the transmission signal structure, the tap coefficient of the equalizer is updated using pilot signals previously promised at the transmitting and receiving sides, and equalized using linear interpolation for data signals between pilot signals.

The channel equalizer 200 according to the present invention is composed of a signal magnitude equalizer 201 and a phase equalizer 202, and the channel equalizer 200 attenuates the signal magnitude for each subcarrier as mentioned above. The absolute value | α _i | and the phase distortion ∠α _i are corrected.

That is, the channel equalizer according to the present invention equalizes signal size attenuation and phase distortion independently as shown in FIG. 3. In Fig. 3, ρ _i and θ _i are the tap coefficients of the signal magnitude equalizer 201 and the phase equalizer 202 at time i, respectively.

4 is the signal amplitude equalizer and a view showing the roles of the phase equalizers, in the case where the pilot signal is modulated from the transmitter to the R + j0 has been received through the channel in y _i signal amplitude equalizer 201 is an equalized signal, Is corrected to R, and the phase equalizer 202 corrects the phase to 0 degrees.

Hereinafter, the signal magnitude equalizer 201 and the phase equalizer 202 will be described in detail.

First, referring to the signal size equalizer 201, the magnitude of the signal equalized by the signal size equalizer 201 The mean squared error between and R is minimized. Therefore, the cost function J _mag is expressed by Equation 5 below.

Tap coefficient of the signal magnitude equalizer 201 in Equation 5 When the slope is taken with respect to Equation 6 below.

In addition, when the highest drop method is used in Equation 5, a tap coefficient update expression as shown in Equation 7 below can be obtained.

Where _mag is a positive constant representing the tap coefficient update size of the signal size equalizer 201.

However, in the tap coefficient update as in Equation 7, Since is always a real number greater than "0", it does not affect the direction of the tap coefficient update, and the signal-to-noise ratio is low If is small, it is a factor to inhibit the tap coefficient update rate.

Therefore, in this case By taking the sign of, the tap coefficient update equation as shown in Equation 8 below is simplified, and the update speed can be increased invariably at the convergence point in the steady state.

Equation 8 is implemented in hardware It is difficult to implement correctly. Therefore, in this case the error function in equation (8) instead It can be done as shown in Equation 9 below.

The signal size equalizer 201 may provide information about the channel size response size required for the Viterbi decoder to be used at the rear of the channel equalizer 200.

Next, the channel equalizer 202 will be described.

Since the signal magnitude equalizer 201 described above does not consider the phase of the received signal, it can be designed independently of the tap coefficient of the signal magnitude equalizer when designing the phase equalizer 202.

In this case, the equalized signal may be represented by Equation 10 below.

Here, θ _i is a tap coefficient of the equalizer and phase.

The phase equalizer aims to make the phase (∠z _i ) of the equalization signal "0". Therefore, maximizing the real component of z _i (Re {z _i }) can equalize the phase.

Therefore, the cost function of the phase equalizer can be expressed as Equation 11 below.

When the inclination of the tap coefficient θ _i of the phase equalizer in Equation 11 is taken, it may be expressed as Equation 12 below.

In addition, if the highest drop method is used in Equation 12, a simple tap coefficient update equation such as Equation 13 below can be obtained.

Where μ _phase is a positive constant representing the tap coefficient update magnitude of the phase equalizer.

However, since the imaginary component Im (z _i ) of the equalization signal has a small value when the phase (∠z _i ) of the equalization signal is close to ± π, a large amount of time is required for the equalizer to converge when Equation 13 is used. I get caught.

Therefore, in this case, by using the sign as shown in Equation 14 below, the convergence point can be quickly reached.

However, when using Equation 14, there is a disadvantage that a residual error occurs after convergence.

Thus, phase In Equation 13, Equation 13 is used, and in other cases Equation 14 is used. This may be expressed as Equation 15 below.

, if Re (z _i ) ≥0

, if Re (z _i ) <0

In the COFDM-based system, the signals received on each subcarrier are sequentially input to the channel equalizer, so that all subchannels can be equalized with one equalizer without having a separate equalizer for each subcarrier. Accordingly, the structure of the channel equalizer in which the above-described signal amplitude and phase equalizer (algorithm) for hardware is implemented will be described with reference to FIG. 5.

FIG. 5 is a block diagram of a channel equalizer of a high definition television according to the present invention. Referring to FIG. 5, a block configuration of an equalized signal, a pilot signal, and a previous tap coefficient value is used. A tap coefficient updater 40 that updates tap coefficients for magnitude and phase equalization, and stores the updated signal magnitude and phase equalized tap coefficient after each updated signal magnitude and phase equalization, respectively. A storage unit 70 outputting the pilot signal position information; The linear interpolator 80 performs linear interpolation using the signal size equalization tap coefficient and the phase equalization tap coefficient respectively stored in the storage unit 70 according to the provided pilot position information, and the phase is corrected in the interpolation unit 80. Linear interpolation unit 80 for the first and second operation units 90 and 91 which take a sine function and a cosine function, respectively, and a result of which the sine and cosine functions are taken by the first and second operation units 90 and 91, respectively. In the first and second multipliers 100 and 110 multiplying the corrected signal size, the complex number processing unit 120 and the complex number processing unit 120 to complex-process the signals output from the first and second multipliers 100 and 110. The multiplier 30 outputs an equalization value by multiplying the processed complex number and the input data, and provides the location information of the pilot signal to the storage unit 70, the tap coefficient updater 40, and the linear interpolator 80, respectively. The control unit 130 is configured. Here, the storage unit 70 is composed of first and second storage units 71 and 72 storing the tap coefficients updated after the signal size and phase equalization, respectively, and the linear interpolation unit 80 is the first and second storage units. The first and second interpolators 81 and 82 linearly interpolate the data signals between the pilot signals provided by using the updated signal magnitude and the phase equalization tap coefficient respectively output from 71 and 72, respectively.

Let us explain the operation of the channel equalizer according to the present invention having such a configuration.

First, the tap coefficient updating unit 40 uses the equalized signal output from the multiplier 30, the pilot signal position information provided from the controller 130, and a previous tap coefficient read from the storage unit 70. Magnitude and phase equalization.

The storage unit 70 stores the signal magnitude and the tap coefficients updated after phase equalization, respectively, in the tap coefficient updater 40, and the linear interpolator 80 linearly interpolates the data signals existing between the pilot signals. Will be taken. When equalization is performed on the pilot signals and data transmitted to all subcarriers as described above, the sine function and the cosine are respectively applied to the output values of the first interpolator 81 to correct the phase through the first and second calculation units 100 and 110. It takes a function.

The sine function and the cosine function are respectively multiplied by the corrected signal size to form a complex number in the complex processor 120 and input to the multiplier 30. The multiplier 30 is a complex value and a multiplier output from the complex processor 120. The equalized output signal is obtained by multiplying the corrected signal output from 110.

Unlike the general LMS method, the channel equalizer of the HDTV according to the present invention as described above equalizes the input signal attenuation and phase distortion independently, thereby providing faster convergence and improved performance in a poor multipath channel environment. Implement channel equalizer to get.

In addition, the complexity is reduced for real hardware implementation, and a structure for equalizing all subchannels sequentially with one channel equalizer is presented. In addition, the signal size equalizer has an advantage of accurately providing information on the channel size response size required for the Viterbi decoder to be used at the rear of the channel equalizer.

Claims (6)

In the channel equalizer of HDTV, A tap coefficient updating unit for updating tap coefficients for magnitude and phase equalization of the signal using the equalized signal, the pilot signal, and previous tap coefficient values; A storage unit for storing the updated signal size and phase equalization tap coefficients, and storing the updated signal magnitude and phase equalization tap coefficients, respectively, and outputting the signal according to the provided pilot signal position information; A linear interpolator for linearly interpolating each of the signal size equalization tap coefficients and phase equalization tap coefficients stored in the storage unit according to the pilot position information provided; First and second calculation units which respectively take a sine function and a cosine function for the phase-corrected signal in the linear interpolator; A complex number processing unit for generating a complex number by multiplying each result of the sine function and the cosine function by the first and second calculation units by the signal magnitude value corrected by the linear interpolation unit; An equalization unit for outputting an equalization value by multiplying the complex number processed by the complex number processing unit and the input data, and a control unit for providing position information of the pilot signal to the storage unit, the tap coefficient update unit, and the linear interpolator, respectively. HDTV channel equalizer. The method of claim 1, And the tap coefficient updating unit updates the tap coefficients using the following equation to equalize the magnitude of the received signal. . Where μ _mag is a positive constant representing the magnitude of the signal tap coefficient update. The method of claim 2, When the signal-to-noise ratio of the received signal is low, the channel equalizer of the HDTV, characterized in that for updating the signal magnitude tap coefficient using the following equation. The method of claim 3, wherein The channel equalizer of the HDTV, characterized in that for updating the signal magnitude tap coefficient by using the following equation to improve the error function to facilitate the hardware when the signal-to-noise ratio of the received signal is low. The method of claim 1, The tap coefficient updating unit has an absolute value of the received signal phase. If larger, the channel equalizer of the HDTV, characterized in that the tap coefficient is updated using the following equation. , Where μ _phase is a positive constant representing the magnitude of the tap coefficient update of the phase equalizer. The method of claim 1, The tap coefficient updating unit has an absolute value of the received signal phase. In the following case, the channel equalizer of the HDTV, characterized in that for updating the tap coefficient using the following equation.