JPH01146448A - Frequency offset correction circuit - Google Patents

Abstract

PURPOSE:To measure a frequency offset without error from an unknown data signal by applying decision through data point arrangement less than the normal data point arrangement and applying correction when the output of an adaptive equalizer is large. CONSTITUTION:Phase difference generating means 5, 6, 7, 8 compare the data point corresponding to the product of the output signal between frequency variable oscillation means 12, 13 applying frequency offset correction and the adaptive equalizer 1 with the data point less than the usual data point for the arrangement and select the data point closest to the data point representing the output signal of a multiplier 4 and generate a phase difference signal between the output signal of the multiplier 4 and the selected data point. Control means 9, 10, 11 give the phase difference signal to frequency variable oscillation means 12, 13 when the absolute value of the output signal of the multiplier 4 is larger than the prescribed value and do not give zero to the frequency variable oscillation means when the absolute value of the output signal of the multiplier is smaller than the prescribed value. Thus, control is not carried out.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is applied to a transmission system for orthogonal amplitude modulated data. In particular, the present invention relates to a frequency offset measuring circuit.

〔overview〕

The present invention provides means for correcting the amount of frequency offset based on the data point arrangement of orthogonal amplitude modulated data, in which a determination is made using a data point arrangement smaller than the normal data point arrangement, and when the output of the adaptive equalizer is large. By making this modification, it is now possible to measure the frequency offset amount from an unknown data signal without error.

[Conventional technology]

Conventionally, this type of frequency offset amount has been estimated based on a data point arrangement for determining a normal received data signal.

[Problem that the invention seeks to solve]

Such a conventional frequency offset correction circuit is effective because it is relatively less likely to make errors in determining data points even if a large amount of code amount interference caused by deterioration due to telephone line characteristics is included when the number of data points is small.

However, when the number of data points is increased (for example, 64 points or 128 points), measurement errors tend to occur and it becomes difficult to estimate the amount of frequency offset from an unknown data signal.

The present invention aims to eliminate such drawbacks, and provides a frequency offset correction circuit that can measure the frequency offset amount from an unknown data signal without error even in quadrature amplitude modulated data with many data points. purpose.

[Means for Solving the Problems] The present invention provides a variable frequency oscillation means, a multiplier that calculates the product of an output signal of the frequency variable oscillation means and an output signal of an adaptive equalizer, and a quadrature amplitude modulated a second data point closest to the first data point corresponding to the output signal of the multiplier among the data points in the data point arrangement having fewer data points than the data points of the data signal; phase difference generating means for generating a phase difference signal indicating a phase difference of; and adjusting the output of the frequency variable oscillation means based on the phase difference signal when the absolute value of the output signal of the multiplier is larger than a predetermined value; It is also characterized by comprising a control means for prohibiting this adjustment when the adjustment is small.

[Effect]

A data point corresponding to the product of the output signal of the frequency variable oscillation means that performs frequency offset correction and the output signal of the adaptive equalizer is compared with a normal data point arrangement having fewer data points, and the output signal of the multiplier is A data point closest to the data point being represented is selected and a phase difference signal between the output signal of the multiplier and the selected data point is generated. If the absolute value of the multiplier output signal is larger than a predetermined value, the phase difference signal is input to the frequency variable oscillation means, and if the absolute value of the multiplier output signal is smaller than the predetermined value, zero is input to the frequency variable oscillation means. control.

〔Example〕

Hereinafter, one embodiment of the present invention will be described based on the drawings. 1st
The figure is a block configuration diagram showing the configuration of this embodiment.

This embodiment, as shown in FIG. a multiplier 4 for calculating the product of the output signal of the means and the output signal of the adaptive equalizer 1; and a multiplier among the data points of the data point arrangement having fewer data points than the data points of the quadrature amplitude modulated data signal. A determination circuit 5 which is a phase difference generating means for generating a phase difference signal indicating a phase difference between this first data point and a second data point nearest to the first data point corresponding to the output signal of No. 4; Adjusting the output of the variable frequency oscillation means based on the phase difference signal when the absolute value of the output signal of the conjugate circuit 6, the multiplier 7, the imaginary part extraction circuit 8, and the multiplier 4 is larger than a predetermined value, and Absolute value converting circuit 9, comparator 10, and switch 11 are control means for prohibiting this adjustment when the value is small.
Equipped with.

The signal received from the transmission channel is input to an adaptive equalizer 1, and a first output of the adaptive equalizer 1 is input to an absolute value squaring circuit 2 which takes the square of the absolute value of its output, and the first output of the adaptive equalizer 1 is input to an absolute value squaring circuit 2 which takes the square of the absolute value of its output. The output of 2 is subtracted to a predetermined value by a subtracter 3, and the subtracted value is input to an adaptive equalizer 1. The second output of the adaptive equalizer 1 is input to one of the two inputs of the multiplier 40, and the output signal of the multiplier 4 is input to the determination circuit 5. The output of the judgment circuit 5 is input to a complex conjugate circuit 6, the output thereof and the output of the multiplier 4 are input to a multiplier 7, the output of the multiplier 7 is input to an imaginary part extraction circuit 8, and the output is input to a switch 11 will be man-powered. The output of the multiplier 4 is input to an absolute value conversion circuit 9, and the output of the absolute value conversion circuit 9 is input to a comparator 10, and a switch 11 is controlled by the output signal. The output of the switch 11 is input to the loop filter 12 , the output of which is input to the variable frequency oscillator 13 , and the output of the variable frequency oscillator 13 becomes the other input of the multiplier 4 .

The adaptive equalization layer 1 requires training, which is typically done with a known data signal sent by the other station. However, in the case of a slave modem such as a multi-point modem, it becomes necessary to train the adaptive equalizer 1 with an unknown data signal. FIG. 3 shows the data point arrangement in the case of 4 points, and FIG. 4 shows the data point arrangement in the case of 128 points. However, as shown in the figure, the received data signal is not clear, and it is difficult to judge it as it is due to code amount interference due to deterioration of the telephone line.
Furthermore, the more data points are arranged, the more difficult the determination becomes. Further, when the adaptive equalizer 1 is to be pulled in, the influence of frequency offset must be removed, but estimation thereof becomes difficult.

The present invention aims to reduce the burden on the adaptive equalizer 1 by estimating the frequency offset as a step before the pull-in of the adaptive equalizer 1 and then letting the adaptive equalizer 1 pull in.

In the present invention, the adaptive equalizer 1 is roughly drawn in using the absolute value square circuit 2 and the subtracter 3 using the technique disclosed in Japanese Patent Publication No. 59-24568, for example, to reduce code amount interference. That is, the correction of the taps of the adaptive equalizer 1 is performed based on the following equation.

Cnil = Ch-α'・X7・Y-(l Y,,l
2R2) Here, Ch is a tap weight vector, x, is an equalization layer input signal vector, Y is an equalization layer output signal, R2 is a constant, n represents time nT, and T is a sampling period.

Although the code amount interference of the output of the adaptive equalizer 1 has been reduced, it is still not sufficient. Therefore, the decision point of the decision circuit 5 is set at the position indicated by the X mark in FIG. 2 to correct the frequency offset. That is, the phase angle between the point determined by the determination circuit 5 using the complex conjugate circuit 6, the multiplier 7, and the imaginary part extraction circuit 8 and the output of the multiplier 4 (ie, the received signal) is determined. Also, if the output of the multiplier 4 is small, the S/N ratio is poor, so the absolute value conversion circuit 9, the comparator 10, and the switch 11 are used to adjust the output of the multiplier 4 to within the circle in FIG. If there is, no correction is made and correction is made when it is outside the circle. The loop filter 12 averages the output signal of the switch 11, the output being a quantity dependent on the frequency offset of the telephone line. The variable frequency oscillator 13 oscillates a frequency based on the value from the loop filter 12. Its output is input to a multiplier 4 which corrects the frequency offset of the output of the adaptive equalizer 1.

Although there are no data points at the locations shown in FIG. 2 in the case of 128 points, the characteristics of the frequency offset measurement are improved over a data point arrangement that is more data point dependent (e.g., has eight data points).

〔Effect of the invention〕

As explained above, the present invention reduces code amount interference by using an adaptive equalizer when training an adaptive equalizer based on an unknown data signal, and arranges data points with fewer data points than the normal data point arrangement. Since the determination is made based on the following and correction is made only when the output of the adaptive equalizer is large, the frequency offset can be measured without error.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram showing the configuration of an embodiment of the present invention. Figure 2 shows the data point arrangement used in the present invention. Figure 3 shows the data point arrangement in the case of 4 points. Figure 4 shows the data point arrangement for 128 points. 1...Adaptive equalizer, 2...Absolute value square circuit, 3...
・Subtractor, 4.7... Multiplier, 5... Judgment circuit, 6
...Complex conjugate circuit, ...Imaginary part extraction circuit, 9...
Absolute value conversion circuit, 10... comparator, 11... switching device,
12... Loop filter, 13... Frequency variable oscillator. Example data point arrangement Fig. 2 1m Data point arrangement (4 points) Fig. 3

Claims (1)

[Claims] (1) In a frequency offset correction circuit connected to an adaptive equalizer which is included in a data receiver to which a quadrature amplitude modulated data signal arrives and is trained based on an unknown data signal, a variable frequency oscillation means (12 , 13), a multiplier (4) for calculating the product of the output signal of the variable frequency oscillation means and the output signal of the adaptive equalizer, and data points smaller than the data points of the orthogonal amplitude modulated data signal. generate a phase difference signal indicating the phase difference between this first data point and a second data point closest to the first data point corresponding to the output signal of the multiplier among the data points of the data point arrangement having Phase difference generating means (5, 6, 7, 8
), and control means (9) that adjusts the output of the variable frequency oscillation means based on the phase difference signal when the absolute value of the output signal of the multiplier is larger than a predetermined value, and prohibits this adjustment when the absolute value is smaller than a predetermined value. , 10, 11).