JP3087627B2 - Demodulator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a demodulator used for digital radio communication, and more particularly to a demodulator having a digital equalizer for equalizing waveform distortion caused by fading or the like.

[0002]

2. Description of the Related Art In recent years, in a digital radio communication system, in order to improve the efficiency of frequency use, multi-level modulation schemes have been promoted. On the other hand, as the number of multivalues increases, the influence of transmission path distortion due to fading or the like increases. In order to remove intersymbol interference and the like caused by the transmission line distortion, waveform equalization by an equalizer is performed. One of the well-known equalizers is a transversal equalizer implemented by digital signal processing in a baseband. The input of this equalizer uses a digital signal that has been subjected to analog-to-digital (A / D) conversion. Before the equalization, the digital signal contains the interference component as it is. On the other hand, in order to perform waveform equalization by an equalizer, it is necessary to perform A / D conversion on a modulated signal to which interference has been added so as not to cause loss of information. That is, the amplitude of the input signal of the A / D converter at the discrimination point must always be within the identifiable range of the A / D converter.

[0003] However, due to the addition of interference,
The amplitude of the input signal of the A / D converter becomes larger than the signal point of the original modulation signal. For example, FIG. 3 shows signal points of an analog baseband signal having four amplitudes. As shown in FIG. 3A, when there is no intersymbol interference, each signal point is naturally within the A / D identification range. However, when intersymbol interference occurs, for example, when an interference component is added to the maximum amplitude signal, the amplitude further increases, and A / A
The identification range of D is exceeded. A signal exceeding the identification range sticks to the maximum value as a digital signal value, and the magnitude of the interference component cannot be correctly recognized. As a result, the waveform equalization capability of the equalizer is greatly deteriorated.

Therefore, in the conventional method, the A / D at the time of steady state
A proposal has been made in which the input signal of the converter is made smaller than the A / D discrimination range at the discrimination point so that the amplitude of the signal does not exceed the discrimination range of the A / D converter even when interference is added. ing. For example, JP-A-63-119331 proposes a demodulator using a baseband transversal type equalizer as shown in FIG.
The modulated wave in the IF band input from the terminal 1 first passes through a variable gain amplifier 11 whose gain changes according to a control signal, and is then detected by a demodulator 12 to output a baseband analog signal. This baseband analog signal is A /
After being converted into a digital signal by the D converter 14, the equalizer 1
6 for waveform equalization. The output signal of the equalizer is multiplied by a coefficient C set in advance by a multiplier 17 and output from a terminal 2.

[0005] For example, in the case of a signal having a quaternary amplitude, the upper two bits of the output signal are the demodulated data signal.
Therefore, the control circuit 18 outputs the output signal of the multiplier 17 and the output signal of FIG.
The deviation of the standard signal point from the reference value shown in FIG.
The gain of the variable gain amplifier is controlled so that the deviation becomes zero. Therefore, the amplitude of the input signal of the A / D converter 14 at the discrimination point is the reciprocal of the coefficient C with respect to the reference value.

For example, when C = 2, the A / D converter 14
The gain of the variable gain amplifier 11 is controlled so that the amplitude at the discrimination point becomes the half signal point shown in FIG. In this case, when there is no intersymbol interference, the A / D converter 14
Uses only half of the identification range. When an interference component is added to the modulated wave, the interference component is also added to the signal point of the analog baseband signal.However, since there is a margin in the identification range, even if interference having the same amplitude is added, the interference range still exceeds the identification range. Never. That is, A / D conversion can be performed without losing information on the interference component.

[0007]

However, in the above-mentioned conventional demodulator, since the amplitude of the input of the A / D converter is reduced, the effective accuracy of the A / D converter is reduced, and the error is accordingly reduced. The rate deteriorates. For example, when the amplitude of the input of the A / D converter is reduced to の of the standard state, the effective accuracy is reduced by one bit, and the quantization noise in the A / D conversion is increased by 6 dB. SUMMARY OF THE INVENTION An object of the present invention is to provide a demodulator capable of effectively using an identification range in an A / D converter to prevent deterioration of error characteristics.

[0008]

A demodulating apparatus according to the present invention comprises a variable gain amplifier for changing the amplitude of an input analog signal based on a control signal, and detecting the amplified signal to output a baseband signal. A demodulator, a compressor for compressing the amplitude of the demodulated signal, an A / D converter for converting the amplitude-compressed signal into a digital signal, and a characteristic having the inverse of the amplitude of the converted digital signal. A decompressor, an equalizer for equalizing an intersymbol interference component included in the decompressed signal, and detecting a difference between an amplitude of an output signal of the equalizer and a reference amplitude, and detecting the difference. And a control circuit for outputting a control signal for controlling the gain of the variable gain amplification so that the value becomes zero.

Here, the compression / expansion characteristics of the compressor and the decompressor are configured as a combination of linear characteristics, and the input / output characteristics of the compressor are such that a signal including an interference component does not exceed the identification range of the A / D converter. When the amplitude of the A / D converter is normalized to ± 1.
It is preferable that the compression ratio outside the above range is made smaller than the compression ratio (compressed amplitude / amplitude before compression) in the range where the input amplitude range is within ± 3/4 , and the compression ratio is constant within the above range . Characterized.

[0010]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of one embodiment of the present invention. Here, for the sake of simplicity, a case of a quaternary amplitude shift keying (ASK) system for transmitting information by a one-dimensional amplitude change (quaternary) will be described. The modulated wave in the IF band input from the terminal 1 passes through the variable gain amplifier 11 and is then detected by the demodulator 12 to output a baseband analog signal. The amplitude of the baseband analog signal is limited by the compressor 13 before the A / D conversion. Although various types of amplitude limiting characteristics are conceivable, basically, the compression ratio outside the signal point range is reduced as compared with the original signal point range of the multilevel signal.

The analog signal compressed by the compressor 13 is A / D converted by an A / D converter 14. A / D
The converted digital signal is input to a decompressor 15 that realizes input / output characteristics opposite to those of the compressor 13 by digital signal processing. Here, the amplitude compressed by the compressor 13 is expanded.

The equalizer 16 receives the decompressed and restored baseband digital signal as input, performs equalization of the intersymbol interference component, and outputs the equalized signal from the terminal 2. For example, 4
In the case of a signal having a value amplitude, the upper two bits of the output signal are the demodulated data signal. The control circuit 18 detects a deviation between the output signal of the multiplier and the reference value of the standard signal point shown in FIG. 4A, and controls the variable gain amplifier 1 so that the deviation becomes zero.
1 is controlled.

Here, the compressor 13 has, for example, input / output compression characteristics shown in FIG. According to this characteristic, when the identification range of the A / D converter 14 in the case of the standard signal point is normalized to ± 1, when the input amplitude range is within ± 3/4, the input and the equal amplitude are output. If it exceeds / 4, the further amplitude is compressed to 1/5 and output. That is, by using this compressor, the A / D converter 1
4 can be expanded to a range of normalized amplitude ± 2.

Further, in the expander 15, FIG.
The input / output expansion characteristics shown in FIG. That is, when the input amplitude range is within ± 3/4, the input and the output are equal in amplitude, and when the input amplitude is over ± 3/4, the amplitude is expanded to five times and output.

Here, the code error characteristic of digital modulation will be considered. When there is no intersymbol interference, errors in transmission symbols are mainly caused by thermal noise. Since it can be assumed that the thermal noise source is white Gaussian noise, symbol errors mainly occur between adjacent symbols. That is, the quantum characteristic of the A / D converter against thermal noise is effective only between adjacent symbols. In the compression characteristic shown in FIG. 2A, the amplitude value (within ± 3/4 of the normalized value) contributing to the discrimination of the value between the symbols is not changed, so that the same code error as in the case of the standard signal point is used. Rate characteristics are obtained. The value of the quantization noise is about half as compared with the demodulator using 1/2 signal point shown in the conventional example.

When intersymbol interference occurs, a transmission symbol error occurs due to intersymbol interference. By performing compression and decompression, the intersymbol interference component of the equalizer input is almost preserved. Since the demodulation device of the present invention having the compression / decompression characteristics shown in FIG. 2 has the same identification range as the conventional demodulation device, it can be expected that the equalization capability is almost equal. However, an amplitude signal exceeding the normalized amplitude ± 3/4 has a greater influence of quantization noise than an amplitude signal having a normalized amplitude within 3/4. Excessive compression increases quantization noise, so it is necessary to determine the compression ratio so as not to affect waveform equalization. To determine the compression ratio, the modulation multi-level number, intersymbol interference amount, A / D
It is necessary to take into account the number of quantization bits of the converter.

FIGS. 3A and 3B show the input / output characteristics of another embodiment of the compressor and the decompressor. In the case of a demodulation device used in a transmission path having a large intersymbol interference amount, for example, within the amplitude range of a quaternary normal signal point (within a normalized amplitude of ± 3/4)
Is a compression ratio of 2/3, and a compression ratio of 1/5 outside the amplitude range of the normal signal point (within ± 3/4 of the normalized amplitude). In this case, A /
The identification range of the D converter can be expanded to the normalized amplitude ± 13/4, and a larger intersymbol interference compensation characteristic can be obtained as compared with the conventional example in which the identification range is the normalized amplitude ± 2. . In this case, the quantization noise value generated by the A / D converter is 3 times smaller than that of the conventional example in which the compression ratio is 1/2.
/ 4.

In the above-described embodiment, the case of the quaternary amplitude shift keying (ASK) system in which information is transmitted by one-dimensional amplitude change (quaternary) has been described. Needless to say, the present invention can be applied to a two-dimensional quadrature multi-level amplitude modulation (QAM) system.

The input / output characteristics of the compressor and the decompressor, in particular, the compression characteristics such as the compression ratio and the decompression ratio can be appropriately set according to the situation. Therefore, the compression characteristics can be provided in an optimal form according to the transmission path, and the identification range of the A / D converter can be effectively used while suppressing an increase in the quantization noise.

[0020]

As described above, the present invention compresses the amplitude of an input analog signal with nonlinear input / output characteristics, converts the amplitude-compressed signal into a digital signal, and converts the signal into a non-linear signal, which is the inverse of a compressor. A system for extending the amplitude of the digital signal converted with the appropriate input / output characteristics is provided, so that the quantization within the original signal point range of the multi-valued signal is not excessively coarsened, and A
By performing the / D conversion, it is possible to suppress the increase in the quantization noise, prevent the error rate characteristic from deteriorating, expand the identification range of the A / D converter, and realize the preferable equalization of the waveform distortion. In the present invention, the identification range of the A / D converter is normalized to ± 1.
Input to reduce the compression ratio of the outside that range than the amplitude range compression ratio within a range within ± 3/4, and since in this range has a compression ratio constant when the compression ratio within the range is constant As compared with the case where the signal points are not equal, a higher bit error rate characteristic can be obtained when all signal points are averaged. Special
When the input amplitude range is within ± 3/4,
Is output, and if it exceeds ± 3/4, the amplitude
By compressing and outputting to 1/5, the A / D converter
The identification range can be extended to the range of normalized amplitude ± 2
You. Also, when using on a transmission line with a large amount of intersymbol interference
Has a compression ratio of 2 / when the input amplitude range is within ± 3/4.
3, and if it exceeds ± 3/4, the compression rate is 1/5.
In this way, the identification range of the A / D converter can be set to the normalized amplitude ± 13 /
4 can be expanded.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a demodulation device of the present invention.

FIG. 2 is a diagram illustrating an example of input / output characteristics of a compressor and a decompressor.

FIG. 3 is a diagram illustrating another example of each input / output characteristic of a compressor and a decompressor.

FIG. 4 shows a standard signal point of a quaternary signal in A / D conversion and 1
It is a figure which shows a / 2 signal point.

FIG. 5 is a block diagram illustrating an example of a conventional demodulation device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Variable gain amplifier 12 Demodulator 13 Compressor 14 A / D converter 15 Decompressor 16 Equalizer 18 Control circuit

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ identification code FI H04L 27/22 H04L 27/22 Z (58) Fields investigated (Int.Cl. ⁷ , DB name) H04L 27/00-27 / 38 H03G 3/20 H03G 3/30 H04B 7/005 H04B 14/04

Claims (4)

(57) [Claims] A variable gain amplifier that changes the amplitude of an input analog signal based on a control signal; a demodulator that detects an amplified signal and outputs a baseband signal;
A compressor for compressing the amplitude of the demodulated signal; an analog-to-digital converter for converting the amplitude-compressed signal to a digital signal; and an expander for expanding the amplitude of the converted digital signal with characteristics opposite to the compressor. , An equalizer for equalizing the intersymbol interference component included in the expanded signal, and detecting a difference between the amplitude of the output signal of the equalizer and a reference amplitude to make the difference zero. A control circuit that outputs a control signal for controlling the gain of the variable gain amplification, the compression and decompression characteristics of the compressor and the decompressor are configured as a combination of linear characteristics, and the input and output characteristics of the compressor are: the signal containing the interference component is a characteristic for compressing the amplitude in the range not exceeding the identification range of the analog-digital converter, and the analog
When the identification range of the digital-to-digital converter is normalized to ± 1
A compression ratio outside the input amplitude range is made smaller than a compression ratio (compressed amplitude / amplitude before compression) in which the input amplitude range is within ± 3/4 , and the compression ratio is within the input amplitude range. A demodulation device characterized by having a constant value. 2. The demodulator according to claim 1, wherein the input signal is a multilevel amplitude shift keying signal or a quadrature multilevel amplitude modulation signal. 3. A set to 1 compression ratio of the input amplitude range of the input amplitude range is within ± 3/4, according to claim 1 or 2, 1/5 compression ratio range exceeding ± 3/4 Demodulator. Wherein the compression ratio of the input amplitude range of the input amplitude range is ± 3/4 within a 2/3 to claim 1 or 2, 1/5 compression ratio range exceeding ± 3/4 The demodulator according to any of the preceding claims.