JP3166913B2 - Automatic equalizer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic equalizer, and more particularly, to an automatic equalizer that compensates for a signal distorted by intersymbol interference and reproduces timing.

[0002]

2. Description of the Related Art In a conventional automatic equalizer, timing recovery is performed by setting insignificant components to the sum of the cutoff error power of a precursor component, the cutoff error power of a postcursor component, and the normalized noise power.

FIG. 4 is a flowchart for explaining a conventional timing reproducing method.

[0004] As shown in FIG. 4, first, the impulse response h _1, h _2, ..., as an input h _L, respective power | h _{1
^{| 2, | h 2 | 2}} , ..., | seek ² | h _L. Next, i =
1, 2, ... L to the component as a significant | a ^2, | h _i
The components that are not significant, the truncation error of the precursor component ^{_{| h 1 | 2, | h}} 2 | 2, ..., | h iM | ^2, a truncation error of post Casa component _{| h i + M + N |} 2 , | H
The sum of _{I + M + N + 1} | ² ,..., | _HL | ² and the normalized noise power _Pn is obtained, and the ratio of a significant component to a nonsignificant component is determined.

[0005] Then, _{i at} which R _i becomes maximum is output as a timing setting signal S _set .

[0006]

However, the value of the significant component is
Is the power of the precursor component | h
_{i-M + 1}|^Two, | H_{i-M + 2}|^Two, ..., | h_i-1|^TwoExists in
The symbol corresponding to this precursor component
When the value of the judgment symbol corresponding to the component
And the difference between the values of the estimated received signals when
Therefore, a determination error is likely to occur.

Accordingly, an object of the present invention is to perform timing reproduction in consideration of characteristic deterioration due to a precursor component having a value close to a significant component, and to perform high-quality signal equalization.

[0008]

In the present invention for solving the above-mentioned problems, first, an impulse response vector h is input and the power | h of each impulse response is input.
^{_{1 | 2, | h 2 |}} 2, ..., | seek ² | h _L. Next, i =
1, 2, ... L to the component as a significant | a ^2, | h _i
Non-significant components are used as precursor components | h
_{i−M + 1} | ² , | h _{i−M + 2} | ² ,..., | h _i−1 | ² and the minimum value of the precursor component truncation error | h
^{_{1 | 2, | h 2 |}} 2, ..., | h iM | ^2, a truncation error of post Casa component ^{_{| h i + M + N |}} 2, | h I + M + N + 1 | 2,
.., | H _L | ² and the normalized noise power P _n , and the ratio of a significant component to a non-significant component is determined. Then, _i that maximizes R _i is output as the timing setting signal S _set .

[0009]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an equivalent circuit according to the present invention. As shown in FIG. 1, an impulse response operation circuit 10
1 receives the received signal _Sr as an input and receives an impulse response vector h = (h ₁ ,
h ₂ ,... h _L ) are obtained and output.

The standardized noise operation circuit 102 receives the received signal S
_{Using r} as an input, a normalized noise power _Pn obtained by normalizing the thermal noise power is obtained and output.

The timing recovery circuit 103 receives the normalized noise power P _n and the impulse response vector h and outputs a timing setting signal S _set .

The filter coefficient output circuit 104 receives the impulse response vector h and the timing setting signal S _set as inputs, and receives a first filter coefficient group T _C1 and a post-filter coefficient group M composed of M elements in the impulse response vector h. A second filter coefficient group T _C2 composed of N pieces corresponding to the cursor component is output.

The counter 105 generates d ^M transmission signal sequences S _CS when there are d types of transmission code levels.

Precursor estimation circuit 106 receives as input first filter coefficient group T _C1 and d ^M transmission signal sequences S _CS , estimates the precursor components of the received signal, and generates d ^M precursor estimation signals S _pr . Output.

Transversal filter 107 receives determination output signal S _d and second filter coefficient group T _C2 as inputs, estimates the post-cursor component of the received signal, and outputs post-cursor estimation signal S _po .

The adder 108 generates a postcursor estimation signal S
_po and d ^M-number of the precursor estimation signal S _{p r} by adding respectively outputs the d ^M pieces of estimated received signals S _er.

The subtractor 109 is input with the received signal S _r and d ^M-number of estimated received signals S _er, performs subtraction, and outputs the d ^M number of estimated error signal _S eerr. Determinator 110 performs reception code determining d ^M number of estimated error signal S _Eerr as input, and outputs to the outside the determination result as the determination output signal S _d.

FIG. 2 is a flowchart showing a timing reproducing method of the timing reproducing circuit 103. As shown in FIG. 2, first, an impulse response vector h is input, and powers | h ₁ | ² , | h _{2 of} respective impulse responses are input.
| ^2, ..., | seek ² | h _L.

Next, for i = 1, 2,... L, the significant component is | h _i | ² , and the insignificant component is | h _{i−M + 1} | ² , | h _{i-M + 2} | ² , ..., | h _i-1
| H ₁ | ² , | h ₂ | ² ,..., | H _iM | ² that are the minimum of | ² and the truncation errors of the precursor components.
| Hi _{+ M + N} | ² , which is the truncation error of the postcursor component
| H _{I + M + N + 1} | ² ,..., | H _L | ² and the normalized noise power P _n
And the ratio of a significant component to a non-significant component is determined.

Then, _{i at} which R _i becomes maximum is output as a timing setting signal S _set .

With the above operation, the timing is reproduced in consideration of the deterioration of the equalization ability due to the precursor component having a value close to the significant component.

FIG. 3 is a graph showing an example of an impulse response of a communication path using the equivalent circuit of the present invention. For example, an automatic equalizer using a first filter coefficient group T _C1 composed of two impulse responses corresponding to precursor components and a second filter coefficient group T _C2 composed of two impulse responses corresponding to postcursor components Is used in a communication path having an impulse response shown in FIG.

Assuming that normalized noise power P _n = 0.01,
The ratio R _i of a significant component to a non-significant component is
For i = 1 to 7,

[0025]

(Equation 1) Becomes

Since R _i takes the maximum value when i = 3,
The timing setting signal S _set = 3, and the first filter coefficient group T _C1 and the second filter coefficient group T _{C2 become}

[0027]

(Equation 2) Becomes

On the other hand, in the conventional method shown in FIG.

[0029]

(Equation 3) And the setting is such that h ₃ having a value close to significant h ₄ exists.

Therefore, according to the present invention, it is possible to perform timing reproduction in consideration of deterioration of the equalization ability due to a precursor component having a value close to a significant component.

[0031]

According to the present invention described above, timing reproduction can be performed in consideration of characteristic deterioration due to a precursor component having a value close to a significant component, and high-quality signal equalization can be realized.

[Brief description of the drawings]

FIG. 1 is a block diagram of an automatic equalizer of the present invention.

FIG. 2 is a flowchart for explaining the operation of the automatic equalizer of the present invention.

FIG. 3 is a graph showing an example of an impulse response of a communication channel.

FIG. 4 is a flowchart for explaining a conventional timing reproduction method;

[Explanation of symbols]

 Reference Signs List 101 impulse response operation circuit 102 normalized noise operation circuit 103 timing reproduction circuit 104 filter coefficient output circuit 105 counter 106 precursor estimation circuit 107 adder 108 transversal filter 109 subtractor 110 determiner

Claims (1)

(57) [Claims] 1. An impulse response operation circuit for obtaining and outputting L impulse responses with a received signal as input, and obtaining and outputting normalized noise power obtained by standardizing thermal noise power with the received signal as input. A normalized noise operation circuit, a timing recovery circuit that receives the normalized noise power and the L number of impulse responses and outputs a timing setting signal, and receives the L number of impulse responses and the timing setting signal as input, A filter coefficient output circuit for outputting a first filter coefficient group consisting of M corresponding to the precursor component and a second filter coefficient group consisting of N corresponding to the postcursor component in the impulse response; a transmission signal sequence generator that generates d ^M transmission signal sequences when the number of types is d; the first filter coefficient group and d ^M A plurality of the transmission signal sequences as inputs, a precursor estimation circuit that estimates a precursor component of the reception signal and outputs d ^M precursor estimation signals, and a determination output signal and the second filter coefficient group as inputs. A transversal filter for estimating a postcursor component of the received signal and outputting a postcursor estimated signal; and adding the postcursor estimated signal and d ^M of the precursor estimated signals to obtain the d ^M estimated received signals. An adder that outputs the received signal and the d ^M estimated received signals, performs subtraction, and outputs d ^M estimated error signals, and outputs the d ^M estimated error signals. A decision unit for performing a reception code decision as an input and outputting a decision result to the outside as the decision output signal; , Of the L impulse responses, a component that is significant with respect to the i-th (i = 1, 2,... L) impulse response is defined as the power of the i-th impulse response, , The minimum of the (i−M + 1) th to (i−1) th impulse responses that are precursor components of the L impulse responses, and the first to first errors that result in a truncation error of the precursor components. The sum of the (i−M) th impulse response power, the (i + M + N) th to Lth impulse response powers that result in a post-cursor component truncation error, and the normalized noise power. An automatic equalizer characterized in that i having the maximum ratio of significant components is selected and output as the timing setting signal.