Digital filter, design method thereof, design device, and digital filter design program

1. A method of providing filter coefficients for a digital filter, comprising:

a first step of generating a first filter coefficient of a symmetric numeric string;

a second step of obtaining a second filter coefficient of the symmetric numeric ring by realizing a second frequency-amplitude characteristic having a contact point at a position where a maximum value is acquired in a first frequency-amplitude characteristic expressed by said first filter coefficient and having a minimum value at said contact point; and

a third step of executing a calculation to obtain a third filter coefficient of the symmetric numeric ring obtained when a first filter having said first filter coefficient and a second filter having said second filter coefficient are cascaded,

wherein said third filter coefficient generated in said third step is determined as the filter coefficient to be obtained, and

wherein, said digital filter multiplies the plurality of filter coefficients with signals at taps of a tapped delay line made up of a plurality of delayers, sums up the multiplication results and outputs the sum.

2. A digital filter which multiplies signals at taps of a tapped delay line made up of a plurality of delayers by given filter coefficients, sums up the multiplication results and outputs the sum,

setting said third filter coefficient determined using the method according to claim 1 to provide the filter coefficients for signals of said respective taps.

3. The method according to claim 1 , wherein the symmetric numeric string is output when a single pulse is input to said first filter and is expressed by {H m , H m−1 , . . . , H 1 , H 0 , H −1 , . . . , H −(m−1) , H −m }, and

said second filter coefficient in said second step is obtained through a calculation {−kH m , −kH m−1 , . . . , −kH 1 , −kH 0 +(1+k), −kH −1 , . . . , −kH −(m−1) , −kH −m }, and

wherein k is an arbitrary positive number.

4. The method according to claim 3 , wherein using said third filter coefficient generated in said third step as said first filter coefficient, the processes in said second step and said third step are repeated two or more times, and

said second filter coefficient is obtained, assuming k≠1, in said second step in a middle stage of said repeated processes, said second filter coefficient is obtained, assuming k=1, in said second step which is a final stage and said third filter coefficient generated in said third step of the final stage is determined as the filter coefficient to be obtained.

5. The method according to claim 1 , wherein

using said third filter coefficient generated in said third step as said first filter coefficient, the processes in said second step and said third step are repeated two or more times to determine the filter coefficient generated in said third step which is the final stage as the filter coefficient to be obtained.

6. The method according to claim 1 , wherein an x-bit (x<y) filter coefficient is obtained by applying rounding of discarding several lower bits to the y-bit data of said third filter coefficient.

7. The method according to claim 6 , wherein the x-bit filter coefficients are subjected to second rounding of multiplying 2 x -fold said x-bit filter coefficients and rounding the decimal fractions so as to round the filter coefficients to an integer.

8. A digital filter design device comprising:

filter coefficient storage means for storing data of a first filter coefficient of a symmetric numeric string; and

calculation means for executing:

a calculation to obtain a second filter coefficient of a symmetric numeric string by realizing a second frequency-amplitude characteristic having a contact point at a position where a maximum value is acquired in a first frequency-amplitude characteristic expressed by said first filter coefficient and having a minimum value at said contact point; and

a calculation to obtain a third filter coefficient obtained when a first filter having said first filter coefficient and a second filter having said second filter coefficient are cascaded,

wherein data of said filter coefficients are stored in said filter coefficient storage means, and

wherein said calculation means uses the data stored in said filter coefficient storage means.

9. A digital filter which multiplies signals at taps of a tapped delay line made up of a plurality of delayers by given filter coefficients, adds up the multiplication results and outputs the addition result, comprising:

an original filter having a first filter coefficient of a symmetric numeric string; and

an adjustment filter having a second filter coefficient of the symmetric string by realizing a frequency-amplitude characteristic having a contact point at a position where a maximum value is acquired in said original filter frequency-amplitude characteristic and having a minimum value at said contact point,

wherein said original filter and said adjustment filter are cascaded.

10. The digital filter according to claim 9 , wherein when the numerical string output from said original filter when a single pulse is input to said original filter is expressed by {H m , H m−1 , . . . , H 1 , H 0 , H −1 , . . . , H −(m−1) , H −m },

said second filter coefficient making up said adjustment filter is {−kH m , −kH m−1 , . . . , −kH 1 , −kH 0 +(1+k), −kH −1 , . . . , −kH −(m−1) , −kH −m } (k is an arbitrary positive number).

11. A digital filter which multiplies signals at taps of a tapped delay line made up of a plurality of delayers by given filter coefficients, adds up the multiplication results and outputs the addition result, comprising:

an original filter having a first filter coefficient of a symmetric numeric string; and

a plurality of adjustment filters having a second filter coefficient of a symmetric numeric string, cascaded in that order,

wherein said second filter coefficients making up said plurality of adjustment filters are symmetric filter coefficients each realizing a frequency-amplitude characteristic having a contact point at a position where a maximum value is acquired in the frequency-amplitude characteristic specified by the numeric string output from an anterior filter when a single pulse is input to said original filter and having a minimum value at said contact point.

12. The digital filter according to claim 11 , wherein when the numeric string output from said anterior filter is assumed to be {H m , H m−1 , . . . , H 1 , H 0 , H −1 , . . . , H −(m−1) , H −m },

a second filter coefficient making up said adjustment filter is {−kH m , −kH m−1 , . . . , −kH 1 , −kH 0 +(1+k), −kH −1 , . . . , −kH −(m−1) , −kH −m } (k is an arbitrary positive number.

13. The digital filter according to claim 12 , comprising said original filter, one or more said adjustment filters in which said second filter coefficients are set as k≠1 and one or more said adjustment filters in which said second filter coefficients are set as k=1, cascaded in that order.

14. A computer-readable storage medium encoded with instructions operable to cause a computer to implement a method of digital filter design, comprising:

generating a first filter coefficient of a symmetric numeric string;

obtaining a second filter coefficient of the symmetric numeric string by realizing a second frequency-amplitude characteristic having a contact point at a position where a maximum value is acquired in a first frequency-amplitude characteristic expressed by said first filter coefficient and having a minimum value at said contact point; and

executing a calculation to obtain a third filter coefficient of the symmetric numeric string obtained when a first filter having said first filter coefficient and a second filter having said second filter coefficient are cascaded,

wherein said third filter coefficient generated in the executing of the calculation is determined as the filter coefficient to be obtained.

15. The digital filter design method according to claim 14 , wherein the symmetric numeric string is output when a single pulse is input to said first filter and is expressed by {H m , H m−1 , . . . , H 1 , H 0 , H −1 , . . . , H −(m−1) , H −m }; and

said second filter coefficient in said second step is obtained through a calculation {−kH m , −kH m−1 , . . . , −kH 1 , −kH 0 +(1+k), −kH −1 , . . . , −kH −(m−1) , −kH −m },

wherein k is an arbitrary positive number.

16. The digital filter design method according to claim 15 , wherein using said third filter coefficient generated in said third step as said first filter coefficient, the processes in said second step and said third step are repeated two or more times, and

said second filter coefficient is obtained assuming k≠1 in said second step in a middle stage of said repeated processes, said second filter coefficient is obtained assuming k=1 in said second step which is the final stage and said third filter coefficient generated in said third step of the final stage is determined as the filter coefficient to be obtained.

17. The digital filter design method according to claim 14 , wherein using said third filter coefficient generated in said third step as said first filter coefficient, the processes in said second step and said third step are repeated two or more times to determine the filter coefficient generated in said third step which is the final stage as the filter coefficient to be obtained.

18. The digital filter design method according to claim 14 , wherein an x-bit (x<y) filter coefficient is obtained by applying rounding of discarding several lower bits to the y-bit data of said third filter coefficient.

19. The digital filter design method according to claim 18 , wherein the x-bit filter coefficients are subjected to second rounding of multiplying 2 x -fold said x-bit filter coefficients and rounding the decimal fractions so as to round the filter coefficients to an integer.