KR100206810B1 - Equalizer and ghost position discriminating method - Google Patents

Abstract

The present invention relates to a symbol mutual interference and ghost cancelable equalizer and a method for determining a ghost position thereof. In the prior art, when symbol ISI and ghost are removed, There is a problem that the removal of the cotton ghost can not be performed. Accordingly, the present invention provides a ghost discrimination unit 200 for discriminating the presence or absence of a ghost from an input signal, determining a position of the ghost in the presence of the ghost and determining an initial tap coefficient to be loaded; The tap updating and equalizing unit 210 is configured to remove symbol mutual interference and ghost using the position of the ghost input from the ghost determining unit and the initial tap coefficient, thereby setting an initial tap coefficient according to the type of ghost, So as to increase the adaptability to various ghosts.

Description

Symbol mutual interference and ghost cancelable equalizer and method of determining ghost position thereof

The present invention relates to a symbol mutual interference and ghost cancelable equalizer for removing a ghost by adjusting a tap coefficient according to a ghost type upon initial tap coefficient loading and a ghost position determining method therefor. And a ghost canceling equalizer and a ghost position determining method for removing a ghost that deviates from a total tap coefficient according to a pre-ghost or a post-ghost.

As shown in FIG. 1, a configuration of an equalizer for a feedback feedback system using a conventional LMS (Least Mean Symbol) algorithm includes a first delay unit 100 for delaying an input ghost signal g _n by a predetermined time, Wow; A first multiplier 110 for multiplying the input ghost signal g _n and a delayed signal through a delay unit by a forward tap coefficient; A first adder 120 for adding output values of respective multipliers of the first multiplier 110; A detection unit 130 for detecting a judgment value of a refund from a signal added through the first adder 120; A second delay unit 140 for delaying the determination value detected through the detector 130 by a predetermined time; A second multiplier 150 for multiplying an output of the second delay unit 140 by an inverse tap coefficient; A second adder 160 for adding the respective multiplier outputs of the second multiplier 150; The output of the first adder 120 is input to the non-inverting terminal (+) and the output of the second adder 160 is input to the inverting terminal (-), And a third interim section 170 for performing a predetermined operation.

The conventional technique configured as described above will be described as follows.

Inter-symbol interference (ISI) has a low-pass filter whose channel characteristics are ideal on one channel, which means mutual interference between symbols due to superposition of symbols due to instability of characteristics.

The ghost is a component generated due to different transmission paths, and has a different time delay for each signal due to different signal transmission paths. That is, the signal transmitted at the same time in the transmitting end means an error component generated due to the mutual interference of these signals because the receiving end receives the signal with a little time error.

In this way, both symbol interference and ghost can be regarded as an error component added to the original signal. These components are to be removed in order to improve the quality of the received signal. These components are eliminated by using an equalizer for the feedback feedback method using the LMS algorithm Let's look at this.

When a signal having the ghost signal gn is input, the first delay unit 100 receives the signal and outputs the delayed signal using a delay T for a predetermined time.

The one output signal is added to take it outputs a forward tap coefficient adjustment to C _M C _-M in the forward tap coefficients in the multipliers 110, and the first adder section 120, the value (U (t)) And outputs it to the third adder 170.

Accordingly, the value (V _n ) added through the third addition unit 170 is output to the detection unit 130, and the detection unit 130 calculates a decision value of the symbols from the added value V _n , And outputs it.

The determination value of the symbols detected through the detection unit 130 in the second delay unit 140 , And the second multiplier 150 multiplies the delay value by the inverse tap coefficients f ₁ to fn and outputs the result.

The values multiplied by the inverse tap coefficients in the second delay unit 140 through the second multiplier 150 are added by the second adder 160 and output to the third adder 170, The output of the first adder 120 received at the terminal (-) and added in the forward direction is input to the non-inverting terminal (+), and the two values are added to remove the symbol interference (ISI) and the ghost.

In other words, the determination value of the detected symbols through the detection unit 130, (ISI) and GOS by eliminating the symbol interference (ISI) by offsetting the past value by an appropriate weight.

This process uses the LMS algorithm, which is an algorithm that adjusts tap coefficients in a direction that minimizes the square of the magnitude of the difference between the input signal and the output signal.

If the tap tap coefficient value is C _n and the tap tap coefficient value is f _n , the error signal e _n is as follows.

[Formula 1]

When the above equation (1) is applied to the LMS algorithm, the tap coefficient is expressed by the following equation.

[Formula 2]

here, Wow Are the step size parameters of the forward and backward equalizers, respectively, which are factors that control convergence speed and stability.

The equalizer operating in this way needs to give an initial value to the tap coefficients at the beginning of the operation. In general, as shown in FIG. 2A, only C _M is set to 1 and the rest is set to all 0s.

The reason for doing this is to detect the error component of the immediately inputted signal at the initial stage of the driving of the equalizer, and the error component at this moment is driven and the error component becomes larger than the signal after the elapsed time.

At this time, the forward and backward tap coefficients are updated using the LMS algorithm to remove the symbol interference (ISI) and ghost on the channel.

Therefore, Cg _n and fa _n in equation (1) can be expressed as follows.

[Formula 3]

In Equation (3), it can be seen that the error component of the LMS algorithm is extracted from the M signals of the future, the current signal, and the past n signals. That is, it eliminates errors between these ranges.

However, in the above conventional art, when ISI and ghost are removed, there is a problem that ghost can not be removed if the total ghost is in the n range and the post-ghost is out of the M range.

SUMMARY OF THE INVENTION It is, therefore, an object of the present invention to solve the above-mentioned problems of the prior art, and an object of the present invention is to provide a method and apparatus for detecting intermittent ghosts and ghosts, And to provide a symbol mutual interference and ghost cancelable equalizer for eliminating ghost that deviates from the entire tap coefficient and a ghost position determining method thereof.

FIG. 1 is a block diagram of an equalizer for a decision feedback system employing a conventional LMS (Least Mean Square) algorithm.

FIG. 2 is an explanatory diagram showing a tap coefficient and a set value in FIG. 1; FIG.

FIG. 3 is a block diagram of an inventive symbol interference (ISI) and ghost cancelable equalizer. FIG.

FIG. 4 is a detailed view of the ghost discriminating portion in FIG. 3; FIG.

FIG. 5 is an input / output signal for obtaining a ground correlation function in FIG. 4; FIG.

FIG. 6 is a flow chart of a method for determining a ghost position of an equalizer capable of mutual interference and ghost cancellation according to the present invention.

DESCRIPTION OF THE REFERENCE NUMERALS

200: ghost discrimination unit 201: autocorrelation function calculation unit

202: threshold value discrimination unit 203: ghost position discrimination unit

210: tap coefficient update and equalization unit

According to an aspect of the present invention, there is provided a method for determining a ghost position of a symbol mutual interference and ghost cancelable equalizer, the method including: a first step of calculating an autocorrelation function from an input signal; A second step of transmitting the autocorrelation function over a threshold in the autocorrelation function calculated in the first step; And a third step of setting an initial value according to the position of the ghost corresponding to the position of the ghost in the portion transferred in the second step to remove the ghost.

As shown in FIG. 3, the symbol mutual interference and ghost cancelable equalizer configuration of the present invention for performing the above-mentioned respective steps determines whether or not a ghost exists from an input signal, A ghost determination unit 200 for determining an initial tap coefficient to be loaded; And a tap coefficient updating and equalizing unit 210 that removes symbol mutual interference and ghost using the position of the ghost input from the ghost determination unit 200 and the initial tap coefficient.

As shown in FIG. 4, the ghost determination unit 200 includes an autocorrelation function calculation unit 201 that calculates and obtains an autocorrelation function from an input signal; A threshold value determiner 202 for transmitting only the portion of the autocorrelation function calculated by the autocorrelation function calculator 201 above a threshold value to the next stage; And a ghost position determining unit 203 for determining the position of the ghost from the signal passed through the threshold determining unit 202 and determining an initial value upon loading.

The operation and effect of the present invention will be described in detail as follows.

When a signal is input to the ghost discriminator 200, the ghost discriminator 200 and the tab update and equalizer 300 receive the signal, the ghost discriminator 200 will be described with reference to FIG.

When a signal is input, the autocorrelation function is calculated by the autocorrelation function calculator 201, and the obtained autocorrelation function is transmitted to the next ghost position determiner 203 only at a portion exceeding the threshold value of the threshold value determiner 202 .

After all, the part beyond the threshold is where ghost exists.

Here, the equation for the autocorrelation function is as follows.

Here, e represents the average of the expressions in parentheses.

If a ghost such as a dotted line exists for an input signal indicated by a solid line as shown in FIG. 5a, the autocorrelation function is output in the same manner as in FIG. 5b.

Then, the threshold value determination unit 202 determines a position of the ghost by comparing a threshold value with a threshold value by adding an appropriate threshold value to the output signal as shown in FIG. 5B.

That is, if the position of the ghost transmitted through the threshold value determiner 202 is between -M and n in FIG. 2, the ghost position determining unit 203 determines the tap coefficient C _M is set to 1, and if it is not within this range, the next step is taken.

Where M and n are the number of defined taps, M is the number of tabs in the forward direction, and n is the number of tabs in the reverse direction.

In the next stage, if the position of the ghost is in a post-ghost state existing between - (M + n) and -M, the initial value is set to 1 and the reverse tap coefficient f _n is set to 1 as shown in FIG. 2 c, In the next stage, if the position of the ghost is n + n, the initial tap coefficient C _-M is set to 1 in the same manner as in FIG.

When the ghost removal range is out of the above range, the forward tap coefficient C _M is set to 1 as an initial value as shown in FIG.

The ghost determination unit 200 outputs the initial value set by the above method to the tap coefficient update and equalizer 210.

Then, the tap coefficient updating and equalizing unit 210 removes the ghost using the initial value transmitted from the ghost determining unit 200. FIG.

As described above, the present invention has an effect of increasing the adaptability to various ghosts by setting initial tap coefficients according to the type of ghost to eliminate various ghosts.

Claims (5)

Ghost discrimination means for discriminating the presence or absence of ghost from an input signal, determining its position in the presence of ghost and determining an initial tap coefficient to be loaded; And a tap updating and equalizing means for removing symbol mutual interference and ghost using the position of the ghost input from the ghost determining means and the initial tap coefficient. The apparatus according to claim 1, wherein the ghost discriminating means comprises: auto-correlation function calculating means for calculating and obtaining an autocorrelation function from an input signal; Threshold value determining means for transmitting only a portion of the autocorrelation function calculated by the autocorrelation function calculating means to a next stage; And ghost position judging means for judging a position of ghost from a signal passed through the threshold value judging means and determining an initial value upon loading. A first step of calculating an autocorrelation function from an input signal; A second step of transmitting the autocorrelation function over a threshold in the autocorrelation function calculated in the first step; And a third step of setting an initial value according to the position of the ghost corresponding to the position of the ghost in the portion transmitted in the second step to remove the ghost. Location identification method. 4. The method of claim 3, wherein in the third step, the initial value setting is set to a forward transform coefficient C _M 1 if the ghost exists between -M and n; A second step of setting the inverse tap coefficient f _n to 1 if the position of the ghost exists between - (M + n) and -M; A third step of setting the forward tap coefficient C- _M to 1 if the position of the ghost exists between n and n + M; And setting a remainder that is set as a tap coefficient 1 to 0 in the process. [5] The method of claim 1, 4. The method of claim 3, further comprising a fourth step of setting an initial value of the forward tap coefficient C _M to 1 if the position of the ghost does not correspond to the ghost removing range in the third step. And a method for determining a ghost position of a ghost cancelable equalizer.