KR0134482B1 - Adaptive squalizer - Google Patents

Abstract

The present invention relates to an adaptive equalizer using an interleaved coefficient update, and includes an adaptive equalizer 10 for filtering and outputting an input signal by filtering coefficients, a reference signal unit 32 and an adaptive equalizer 10 for outputting a reference signal. ) Output buffer 31 for storing the filtered output signal for a predetermined time and outputting the output signal according to the coefficient update period (K) and outputting these signals to the comparator 50 according to the synchronization control signal of the input signal. Comparing unit for comparing the error calculation logic 30, the output buffer 31 of the error calculation logic 30 and the output signal of the reference signal unit 32, and then detects the error signal and outputs it to the adaptive logic 40 (50) and the error signal input from the comparator 50 are updated according to the coefficient update period K, and the updated coefficient value is provided to the adaptive equalizer 10 to equalize by the updated coefficient. Configured with adaptation logic 40 to make .

Therefore, since the present invention updates coefficients according to a predetermined predetermined update period, hardware can be easily and efficiently implemented in real time, and more complicated algorithms can be applied. Since the coefficients are updated once, the idle time can be reduced to shorten the information transmission time.

Description

Adaptive Equalizer Using Interleaved Coefficient Update

1 is a block diagram of an adaptive equalizer using a conventional coefficient update.

2 is an input data format of an adaptive equalizer using interleaved coefficient update according to the present invention.

3 is a block diagram of an adaptive equalizer using interleaved coefficient update according to the present invention.

FIG. 4 is a diagram showing an embodiment of filtering and coefficient updating in the case of performing coefficient update by setting the update period to 3. FIG.

5 is a block diagram showing in detail the coefficient update by the update period according to the present invention.

* Explanation of symbols for main parts of the drawings

1, 10: adaptive equalizer 2, 20: decision unit

3: coefficient update unit 4, 32: reference signal unit

5, 50: comparison unit 30: error calculation logic

31: output buffer 40: adaptive logic

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an adaptive equalizer, and more particularly, to an adaptive equalizer using an INTERLEAVED coefficient update that performs coefficient update once every predetermined number of samples without performing coefficient update for every sample of input data.

The signal transmitted from the transmitter is distorted due to deformation by Gaussian thermal noise, impulsive noise, addition or multiplication noise by fading, temporal dispersion, and the like through the transmission channel.

Such a distortion may appear as a deterioration of image quality only in a conventional analog TV system, but in a digital transmission HDTV system, a bit detection error may occur during signal processing at a receiver, making it impossible to restore the entire screen or display a completely different image. . In particular, multiple paths due to time delay and phase change of a transmission signal cause severe intersymbol interference, which is a major cause of bit detection error. This technique of reducing bit detection error at the receiver by compensating for distortion caused by non-ideal transmission channels is called channel equalization.

On the other hand, in such channel equalization, a method of performing coefficient update on an input data sample is widely used. In general, in an adaptive equalizer system that performs equalization using coefficient update, coefficient update is performed every symbol rate (SYMBOL RATE). It is configured to run.

1 is a block diagram of an equalizer using a conventional conventional coefficient update.

As shown in FIG. 1, an adaptive equalizer for performing channel equalization using conventional coefficient updating is an adaptive equalizing unit 1, a determining unit 2, a coefficient updating unit 3, and a reference signal unit. (4) and comparison unit 5 are included.

Looking at the operation of the conventional adaptive equalizer configured as described above are as follows.

First, when a received signal is input to the adaptive equalizer 1, the adaptive equalizer 1 filters the input signal by a predetermined filtering coefficient, and filters the filtered signal by the determination unit 2 and the comparison unit ( Output to 5). The comparator 5 detects an error signal by comparing the filtered signal provided from the adaptive equalizer 1 with the reference signal provided from the reference signal section 4. That is, the comparison unit 5 compares the output signal filtered by the adaptive equalizer 1 with the reference signal output from the reference signal unit 4, detects an error signal, and provides the coefficient update unit 3 with each sample. do. Then, the coefficient updater 3 updates the filter coefficients by executing coefficient update at every symbol rate according to the error signal provided from the comparator 5, and provides the updated coefficient value to the adaptive equalizer 1 again. To filter the input signal by the updated coefficients.

However, as described above, performing coefficient update at every symbol rate is not only very difficult and inefficient to implement the hardware in accordance with the symbol period of the current HDTV system, but even more complicated algorithms are difficult to apply even if implemented.

Accordingly, the present invention has been made to solve the problems described above, and an object of the present invention is to provide an interleaved coefficient that updates coefficients once every few samples without performing coefficient updates every sample in an adaptive equalizer having training data. It is to provide an adaptive equalizer using update.

In order to achieve the above object, the present invention provides a filtering coefficient based on a distortion value of a frequency characteristic caused by transmission through a system or a channel based on a filtered value for the input signal and an error value based on a preset reference signal. An adaptive equalizer for updating, comprising: an adaptive equalizer for equalizing an input signal by filtering an input signal composed of training data consisting of M symbols and an information symbol consisting of N symbols according to a predetermined filtering coefficient; It consists of a reference signal unit for outputting a predetermined reference signal and an output buffer for temporarily storing only the filtered output of the training data of the output signal filtered by the adaptive equalizing unit for a predetermined time and then outputs the output based on a preset coefficient update period. An error calculation logic for outputting the reference signal and the filtered training data according to the synchronization control signal in the input signal; A comparison unit for detecting an error signal by comparing an output buffer of the error calculation logic and an output signal of a reference signal unit; Adaptive logic for performing coefficient update according to the predetermined coefficient update period and providing the updated coefficient value to the adaptive equalizer so that equalization is performed according to the updated coefficient value. Include.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

2 shows a format of input data input to an adaptive equalizer using interleaved coefficient update according to the present invention.

As shown in FIG. 2, the training data of the M symbol of the header part is located in the input data, and the information symbol of the N symbol is located in the rest of the input data.

3 is a block diagram of an adaptive equalizer using interleaved coefficient update according to the present invention.

As shown in FIG. 3, the present invention provides an adaptive equalizer 10 that equalizes an input signal by filtering an input signal according to a predetermined filtering coefficient, and an output signal filtered by the adaptive equalizer 10 for a predetermined time. An output buffer 31 for temporarily storing a reference signal and a reference signal unit 32 for outputting a predetermined reference signal, and outputting a reference signal and a filtered signal to the comparator 50 according to a synchronous control signal of the input signal. Comparing unit for comparing the output logic of the calculation logic 30 and the output buffer 31 and the reference signal unit 32 of the error calculation logic 30 to detect an error signal and output it to the adaptive logic ( 50 and the adaptation logic 40 and the decision unit for performing the coefficient update according to the error signal input from the comparator 50 and providing the updated coefficient value to the adaptive equalizer 10 so that equalization is performed. 20). Here, the decision unit 20 executes a normal decision function.

In a general HDTV system (DSC-HDTV), a training signal of 668 symbols is transmitted in one field. According to the present invention, in order to perform equalization in an adaptive equalizer having such a training signal, all filters until the next training signal is received. All you need to do is update the coefficients for the FILTER TAP.

FIG. 2 shows the format of the input data. As shown in FIG. 2, the input data format has a much larger amount of information symbols than the training signal, and if the coefficient is updated by the training signal symbol until the next training signal comes, do. Therefore, the present invention is configured to set an update period in units of a predetermined number of samples and to execute coefficient update once per set update period.

Also, as shown in FIG. 2, if the number of symbols of the training signal is M, the coefficients to be found are C (n + 1), C (n + 2),... M coefficients must be updated with C (n + M). The time limit for updating M coefficients is N + M symbol periods, where N represents the number of information symbols.

Referring to FIG. 3 again, the reference signal unit 32 provides a preset reference signal to the comparison unit 50 in synchronization with the output buffer 31. The output buffer 31 temporarily stores only the filtered output of the training signal among the filtered signals output from the adaptive equalizing unit 10, and then, the output buffer 31 is synchronized with the reference signal at a preset update period K and to the comparison unit 50. to provide. That is, the error calculation logic 30 is provided to the comparison unit 50 in synchronization with the filtered training signal and the reference signal according to the preset update period K.

Then, the comparison unit 50 compares the filtered training signal with a reference signal, detects an error signal according to the difference value, and provides the detected error signal to the adaptation logic 40.

Accordingly, the adaptation logic 40 executes the coefficient update according to the error signal provided from the comparator 50 at the preset update period K and provides the updated coefficient to the adaptive equalizer 10. Then, the adaptive equalizer 10 performs channel equalization on the input signal by performing filtering according to the updated coefficient.

4 shows coefficient update when the coefficient update period K is 3, and shows that one coefficient update is performed every three sample periods. In addition, Fig. 5 shows the coefficient update according to the present invention in more detail. When coefficient update is performed as shown in Fig. 5, all coefficients can be updated until the next training signal.

As described above, the present invention updates the coefficients once every few samples, that is, according to the update period K, so that hardware can be easily and efficiently implemented in real time, and can be applied to more complex algorithms. By reducing the (IDLE) section, the information transmission time can be shortened.

Claims (1)

An adaptive equalizer that updates the filtering coefficients based on a filtered value for the input signal and an error value based on a predetermined reference signal, for distortion of the frequency characteristic caused by transmission over a system or channel, with M symbols. An adaptive equalizer 10 for equalizing the input signal by filtering the input signal composed of the training data and the information symbols consisting of N symbols according to a predetermined filtering coefficient; Only the filtered output of the training data among the output signals filtered by the reference signal unit 32 and the adaptive equalizer 10 that output a preset reference signal is temporarily stored for a predetermined time and then output based on a preset coefficient update period. An error calculation logic (30) configured to output the reference signal and filtered training data according to a synchronous control signal among the input signals; A comparison unit 50 for comparing the output buffer 31 of the error calculation logic 30 and the output signal of the reference signal unit 32 to detect and output an error signal; The coefficient signal is updated according to the preset coefficient update period K by the error signal provided from the comparator 50, and the updated coefficient value is provided to the adaptive equalizer 10 according to the updated coefficient value. Adaptive equalizer using interleaved coefficient update including adaptation logic 40 to allow equalization.