US 6,888,479 B2 · Data processing method, circuit, and apparatus with increased accuracy

A data processing method of converting (n×m−1)-bit data to multi-level data of m symbols of n bits per symbol, n and m being integers satisfying n≧2 and m≧2, respectively, wherein:

the m symbols each including n-bit data are arranged in m columns in an n×m matrix, with a value of each of the m symbols being set to an even or odd number, by creating the m symbols each including (n−1)-bit data of the (n×m−1)-bit data, arranging the m symbols in the m columns so that the {(n−1)×m}-bit data is arranged in upper-side {(n−1)×m} bits of the n×m matrix, and converting the remaining (m−1)-bit data of the (n×m−1)-bit data to m-bit data so that the m-bit data is arranged in a row in lower-side m bits of the n×m matrix.

A data processing method of determining a value of each of m symbols of a set of a multi-level signal based on input signal data of the m symbols obtained after quantizing the multi-level signal recorded on an information recording medium or transmitted to a transmission path and reproduced from the information recording medium or received from the transmission path, in which multi-level signal each of the m symbols is formed of n-bit data, the set is formed of the m symbols, {(n−1)×m}-bit data is arranged in upper-side (n−1) bits of the m symbols, (m−1)-bit data is converted to m-bit data, and the value of each of the m symbols is set to an even or odd number by arranging the m-bit data in the remaining lower-side bits of the m symbols, n and m being integers satisfying n≧2 and m≧2, respectively, wherein:

2(m−1) ideal signal data value strings corresponding to symbol value strings of even and odd symbol values closest to input signal data values of the m symbols of the set are determined, the symbol value strings corresponding to 2(m−1) m-bit data patterns of m even and odd numbers based on one of which data patterns the (m−1)-bit data is converted to the m-bit data;
a sum of errors between the input signal data values and ideal signal data values of each of the ideal signal data value strings is calculated;
and
one of the symbol value strings is determined to be multi-level data of the set, the one of the symbol value strings corresponding to one, of the ideal signal value strings which one has the minimum sum of errors.

1 more element

A data processing method of determining a value of each of m symbols of a set of a multi-level signal based on input signal data of the m symbols obtained after quantizing the multi-level signal recorded on an information recording medium or transmitted to a transmission path and reproduced from the information recording medium or received from the transmission path, in which multi-level signal:

each of the m symbols is formed of n-bit data;
the set is formed of the m symbols;
{(n−1)×m}-bit data is arranged in upper-side (n−1) bits of the m symbols;
(m−1)-bit data is converted to m-bit data based on one of first and second tables of different types each created to store 2(m−1) m-bit data patterns selected from 2m m-bit data patterns, the first table being selected in the case of a numeric string for selecting the one of the first and second tables for a given set=0 and the second table being selected in the case of the numeric string=1, the numeric string being started from an initial value corresponding to a first set and using a result of a logical operation on a value of the numeric string for an (i−1)th set and predetermined data in the (i−1)th set as a value of the predetermined numeric string for an ith set, i being an integer satisfying i≧2;
and the value of each of the m symbols is set to an even or odd number by arranging the m-bit data in the remaining lower-side bits of the m symbols, n and m being integers satisfying n≧2 and m≧2, respectively, wherein:
2(m−1) ideal signal data value strings corresponding to symbol value strings of even and odd symbol values closest to input signal data values of the m symbols of each of a target set and one or a plurality of succeeding sets in each of cases of the numeric string=0 and 1 are determined, the symbol value strings corresponding to patterns of even and odd numbers on the selected one of the first and second tables;
a sum of errors between the input signal data values and ideal signal data values of each of the ideal signal data value strings is calculated with respect to the m symbols of each of the target and the one or plurality of succeeding sets;
one of the symbol value strings which one corresponds to one of the ideal signal value strings which one has the minimum sum of errors is determined, with respect to each of the target and the one or plurality of succeeding sets in each of the cases of the numeric string=0 and 1, to be temporary multi-level data of each of the target and the one or plurality of succeeding sets;
and
one of the symbol value strings of the target set which one corresponds to one of combinations of the temporary multi-level data of the target set and the one or plurality of succeeding sets which one minimizes a total of the sums of errors of the temporary multi-level data of the target set and the one or plurality of the succeeding sets is determined to be multi-level data of the target set, the combinations being made in accordance with the numeric string.

7 more elements

A signal processing circuit comprising:

(n−1) first selector circuits each supplied with different m-bit data of (n×m−1)-bit parallel data and outputting one-bit data of the supplied m-bit data, n and m being integers satisfying n≧2 and m≧2, respectively;
a converter circuit supplied with the remaining (m−1)-bit data of the (n×m−1)-bit parallel data and converting the supplied (m−1)-bit data to m-bit data;
a second selector circuit supplied with the m-bit data output from said converter circuit and outputting one-bit data of the supplied m-bit data;
a first control circuit outputting the (n×m−1)-bit parallel data as n-bit parallel data;
an output circuit supplied with signal data values obtained after quantizing a multi-level signal whose one symbol is formed of n-bit data and outputting ideal signal data values corresponding to even and odd symbol values closest to each of the supplied signal data values;
a first calculator circuit calculating errors between the supplied signal data values and the respective ideal signal data values;
a plurality of first register circuits retaining the errors calculated in said first calculator circuit;
a second calculator circuit calculating a sum of errors of each of predetermined combinations of the errors output from said first register circuits;
a plurality of second register circuits retaining the sums calculated in said second calculator circuit;
a minimum value detector circuit detecting a minimum value of the sums output from said second register circuits;
and
a second control circuit outputting a symbol value string that is n-bit parallel data, the symbol value string corresponding to the minimum value detected in said minimum value detector circuit.

9 more elements

A data processing method of determining a value of each of m symbols of a set of a multi-level signal based on input signal data of the m symbols obtained after quantizing the multi-level signal recorded on an information recording medium or transmitted to a transmission path and reproduced from the information recording medium or received from the transmission path, in which multi-level signal each of the m symbols is formed of n-bit data, the set is formed of the m symbols, {(n−k)×m}-bit data is arranged in upper-side (n−k) bits of the m symbols, and (m×k−1)-bit data is converted to (m×k)-bit data to be arranged in the remaining lower-side k bits of the m symbols, n, m, and k being integers satisfying n≧2, m≧2, and n>k≧1, respectively, wherein:

2(m×k−1) ideal signal data value strings corresponding to symbol value strings of even and odd symbol values closest to input signal data values of the m symbols of the set are determined, the symbol value strings corresponding to 2(m×k−1) (m×k)-bit data patterns of m even and odd numbers based on one of which data patterns the (m×k−1)-bit data is converted to the (m×k)-bit data;
a sum of errors between the input signal data values and ideal signal data values of each of the ideal signal data value strings is calculated;
and
one of the symbol value strings is determined to be multi-level data of the set, the one of the symbol value strings corresponding to one of the ideal signal value strings which one has the minimum sum of errors.

1 more element

A data processing method of determining a value of each of m symbols of a set of a multi-level signal based on input signal data of the m symbols obtained after quantizing the multi-level signal recorded on an information recording medium or transmitted to a transmission path and reproduced from the information recording medium or received from the transmission path, in which multi-level signal:

each of the m symbols is formed of n-bit data;
the set is formed of the m symbols;
{(n−k)×m}-bit data is arranged in upper-side (n−k) bits of the m symbols;
(m×k−1)-bit data is converted to (m×k)-bit data based on one of first and second tables of different types each created to store 2(m×k−1) (m×k)-bit data patterns selected from 2(m×k) (m×k)-bit data patterns, the first table being selected in the case of a numeric string for selecting the one of the first and second tables for a given set=0 and the second table being selected in the case of the numeric string=1, the numeric string being started from an initial value corresponding to a first set and using a result of a logical operation on a value of the numeric string for an (i−1)th set and predetermined data in the (i−1)th set as a value of the predetermined numeric string for an ith set, i being an integer satisfying i≧2;
and the (m×k)-bit data is arranged in the remaining lower-side k bits of the m symbols, n, m, and k being integers satisfying n≧2, m≧2, and n>k≧1, respectively, wherein:
2(m×k−1) ideal signal data value strings corresponding to symbol value strings of even and odd symbol values closest to input signal data values of the m symbols of each of a target set and one or a plurality of succeeding sets in each of cases of the numeric string=0 and 1 are determined, the symbol value strings corresponding to patterns of even and odd numbers on the selected one of the first and second tables;
a sum of errors between the input signal data values and ideal signal data values of each of the ideal signal data value strings is calculated with respect to the m symbols of each of the target and the one or plurality of succeeding sets;
one of the symbol value strings which one corresponds to one of the ideal signal value strings which one has the minimum sum of errors is determined, with respect to each of the target and the one or plurality of succeeding sets in each of the cases of the numeric string=0 and 1, to be temporary multi-level data of each of the target and the one or plurality of succeeding sets;
and
one of the symbol value strings of the target set which one corresponds to one of combinations of the temporary multi-level data of the target set and the one or plurality of succeeding sets which one minimizes a total of the sums of errors of the temporary multi-level data of the target set and the one or plurality of the succeeding sets is determined to be multi-level data of the target set, the combinations being made in accordance with the numeric string.

7 more elements

A signal processing circuit comprising:

(n−k) first selector circuits each supplied with different m-bit data of (n×m−1)-bit parallel data and outputting one-bit data of the supplied m-bit data, n, m, and k being integers satisfying n≧2, m≧2, and n>k≧1, respectively;
a converter circuit supplied with the remaining (m×k−1)-bit data of the (n×m−1)-bit parallel data and converting the supplied (m×k−1)-bit data to (m×k)-bit data;
and
a second selector circuit supplied with the (m×k)-bit data output from said converter circuit and outputs one-bit data of the supplied (m×k)-bit data;
a first control circuit outputting the (n×m−1)-bit parallel data as n-bit parallel data;
an output circuit supplied with signal data values obtained after quantizing a multi-level signal whose one symbol is formed of n-bit data and outputting ideal signal data values corresponding to a plurality of symbol values closest to each of the supplied signal data values;
a first calculator circuit calculating errors between the supplied signal data values and the respective ideal signal data values;
a plurality of first register circuits retaining the errors calculated in said first calculator circuit;
a second calculator circuit calculating a sum of errors of each of predetermined combinations of the errors output from said first register circuits;
a plurality of second register circuits retaining the sums calculated in said second calculator circuit;
a minimum value detector circuit detecting a minimum value of the sums output from said second register circuits;
and
a second control circuit outputting a symbol value string that is n-bit parallel data, the symbol value string corresponding to the minimum value detected in said minimum value detector circuit.

10 more elements

A multi-level data processing method for converting binary data to multi-level data of n bits per symbol and recording the multi-level data and test data on an information recording medium, the test data including 2(M×n) combinations of M successive multi-level data, n and M being integers satisfying n≧2 and M≧3, respectively, wherein:

one set is formed of m symbols of multi-level data, m being an integer satisfying m≧2;
{(n−k)×m}-bit binary data of (n×m−1)-bit binary data is arranged in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
and
the remaining (m×k−1)-bit binary data of the (n×m−1)-bit binary data is converted to (m×k) bits to be arranged in the remaining lower-side k bits of the m symbols of the set,
whereby the binary data is converted to the multi-level data.

2 more elements

A multi-level data processing method for converting binary data to multi-level data of n bits per symbol, recording the multi-level data and test data on an information recording medium, and reproducing the multi-level data from the information recording medium, the test data including 2(M×n) combinations of M successive multi-level data, n and M being integers satisfying n≧2 and M≧3, respectively, wherein:

at a time of recording the multi-level data, one set is formed of m symbols of multi-level data, m being an integer satisfying m≧2;
{(n−k)×m}-bit binary data of (n×m−1)-bit binary data is arranged in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
and the remaining (m×k−1)-bit binary data of the (n×m−1)-bit binary data is converted to (m×k) bits to be arranged in the remaining lower-side k bits of the m symbols of the set, whereby the binary data is converted to the multi-level data;
and
at a time of reproducing the multi-level data, signal values of multi-level data are stored in a table by inputting a reproduction signal of the test data;
the multi-level data to which the binary data is converted is reproduced to be input as a signal so that first multi-level data minimizing an error between signal values of the input multi-level data and corresponding signal values stored in the table is output;
lower-side k bits of m symbols of a target set of the first multi-level data are compared with conversion data for converting the (m×k−1)-bit binary data to the (m×k) bits;
and one of candidate multi-level data corresponding to the conversion data which one has signal values having a minimum error with respect to the signal values of the input multi-level data is output as second multi-level data and the second multi-level data is output as reproduced multi-level data when the lower-side k bits match none of the conversion data, and the first multi-level data is output as the reproduced multi-level data when the lower-side k bits match any of the conversion data.

5 more elements

A multi-level data processing method for converting binary data to multi-level data of n bits per symbol, recording the multi-level data and test data on an information recording medium, and reproducing the multi-level data from the information recording medium, the test data including 2(M×n) combinations of M successive multi-level data, n and M being integers satisfying n≧2 and M≧3, respectively, wherein:

at a time of recording the multi-level data, one set is formed of m symbols of multi-level data, m being an integer satisfying m≧2;
{(n−k)×m}-bit binary data of (n×m−1)-bit binary data is arranged in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
and the remaining (m×k−1)-bit binary data of the (n×m−1)-bit binary data is converted to (m×k) bits to be arranged in the remaining lower-side k bits of the m symbols of the set, whereby the binary data is converted to the multi-level data;
and
at a time of reproducing the multi-level data, signal values of multi-level data are stored in a table by inputting a reproduction signal of the test data;
the multi-level data to which the binary data is converted is reproduced to be input as a signal so that first multi-level data minimizing an error between signal values of the input multi-level data and corresponding signal values stored in the table is output;
lower-side k bits of m symbols of a target set of the first multi-level data are compared with conversion data for converting the (m×k−1)-bit binary data to the (m×k) bits;
and one of strings of candidate multi-level data of a plurality of successive sets including the target set is selected, the one of the strings having signal values having a minimum error with respect to input signal values of the sets, the candidate multi-level data corresponding to the conversion data, and candidate multi-level data of the target set in the selected one of the strings is output as second multi-level data of the target set so that the second multi-level data is output as reproduced multi-level data when the lower-side k bits match none of the conversion data, and the first multi-level data is output as the reproduced multi-level data when the lower-side k bits match any of the conversion data.

5 more elements

A multi-level data processing apparatus comprising:

a conversion part converting input binary data to multi-level data of n bits per symbol, n being an integer satisfying n≧2;
a generation part generating test data including 2(M×n) combinations of M successive multi-level data, M being an integer satisfying M≧3;
and
a recording part recording the multi-level data output from said conversion part and the test data output from said generation part on an information recording medium,
wherein said conversion part converts input (n×m−1)-bit binary data to multi-level data of m symbols per set, m being an integer satisfying m≧2,
said conversion part comprising:
a first part arranging {(n−k)×m}-bit binary data of the (n×m−1)-bit binary data in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
a second part converting the remaining (m×k−1)-bit binary data of the (m×k−1)-bit data to (m×k) bits;
and
a third part arranging the (m×k) bits in the remaining lower-side k bits of the m symbols of the set.

7 more elements

A multi-level data processing apparatus comprising:

a conversion part converting input binary data to multi-level data of n bits per symbol, n being an integer satisfying n≧2;
a generation part generating test data including 2(M×n) combinations of M successive multi-level data, M being an integer satisfying M≧3;
a recording part recording the multi-level data output from said conversion part and the test data output from said generation part on an information recording medium;
and
a reproduction part reproducing the multi-level data from the information recording medium,
wherein:
said conversion part converts input (n×m−1)-bit binary data to multi-level data of m symbols per set, m being an integer satisfying m≧2,
said conversion part comprising:
a first part arranging {(n−k)×m}-bit binary data of the (n×m−1)-bit binary data in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
a second part converting the remaining (m×k−1)-bit binary data of the (m×k−1)-bit data to (m×k) bits;
and
a third part arranging the (m×k) bits in the remaining lower-side k bits of the m symbols of the set;
and
said reproduction part comprises:
a reproduction signal output part outputting a reproduction signal from the information recording medium;
a signal value output part supplied with the reproduction signal output from said reproduction signal output part and outputting signal values of multi-level data;
a storage part storing the signal values of the test data output from said signal value output part in a table;
a first error calculation part calculating an error between signal values of the multi-level data to which the binary data is converted and corresponding signal values stored in the table;
a first multi-level data output part outputting first multi-level data minimizing the error output from said first error calculation part;
a comparison part comparing lower-side k bits of m symbols of a target set of the first multi-level data with conversion data for converting the (m×k−1)-bit binary data to the (m×k) bits;
a second error calculation part calculating, based on the comparison of said comparison part, an error between each of candidate multi-level data corresponding to the conversion data and the signal values of the multi-level data to which the binary data is converted,
a second multi-level data output part outputting one of the candidate multi-level data which one has a minimum of the errors output from said second error calculation part as second multi-level data;
and
a reproduced multi-level data output part outputting, as reproduced multi-level data, the second multi-level data if the lower-side k bits match none of the conversion data in said comparison part and the first multi-level data if the lower-side k bits match any of the conversion data in said comparison part.

21 more elements

A multi-level data processing apparatus comprising:

a conversion part converting input binary data to multi-level data of n bits per symbol, n being an integer satisfying n≧2;
a generation part generating test data including 2(M×n) combinations of M successive multi-level data, M being an integer satisfying M≧3,
a recording part recording the multi-level data output from said conversion part and the test data output from said generation part on an information recording medium;
and
a reproduction part reproducing the multi-level data from the information recording medium,
wherein:
said conversion part converts input (n×m−1)-bit binary data to multi-level data of m symbols per set, m being an integer satisfying m≧2,
said conversion part comprising:
a first part arranging {(n−k)×m}-bit binary data of the (n×m−1)-bit binary data in upper-side (n−k) bits of m symbols of a set, k being an integer satisfying n>k≧1;
a second part converting the remaining (m×k−1)-bit binary data of the (m×k−1)-bit data to (m×k) bits;
and
a third part arranging the (m×k) bits in the remaining lower-side k bits of the m symbols of the set;
and
said reproduction part comprises:
a reproduction signal output part outputting a reproduction signal from the information recording medium;
a signal value output part supplied with the reproduction signal output from said reproduction signal output part and outputting signal values of multi-level data;
a storage part storing the signal values of the test data output from said signal value output part in a table;
a first error calculation part calculating an error between signal values of the multi-level data to which the binary data is converted and corresponding signal values stored in the table;
a first multi-level data output part outputting first multi-level data minimizing the error output from said first error calculation part;
a comparison part comparing lower-side k bits of m symbols of a target set of the first multi-level data with conversion data for converting the (m×k−1)-bit binary data to the (m×k) bits;
a second error calculation part calculating, based on the comparison in said comparison part, an error between input signal values of the sets and each of strings of candidate multi-level data of a plurality of successive sets including the target set, the candidate multi-level data corresponding to the conversion data;
a second multi-level data output part outputting, as second multi-level data of the target set, candidate multi-level data of the target set in one of the strings of the candidate multi-level data which one has a minimum of the errors calculated in said second error calculation part;
and
a reproduced multi-level data output part outputting, as reproduced multi-level data, the second multi-level data when the lower-side k bits match none of the conversion data, and the first multi-level data when the lower-side k bits match any of the conversion data.

21 more elements

Data processing method, circuit, and apparatus with increased accuracy

Life of this patent

Data processing method, circuit, and apparatus with increased accuracy

Life of this patent

Abstract

Claims79