JPH11274940A - Reed-solomon decoding method using modified berlekamp-massey algorithm, and decoder thereof - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a Reed-Solomon decoding method using a modified Berlekamp- Massey algorithm that is a compromise between the Berlekamp-Massey algorithm and a Euclidian algorithm and to provide its decoder. SOLUTION: This Reed-Solomon decoder provided with a timing and control signal generating section 71, an error location polynomial calibration term register section 72, and error location register section 75, a syndrome register section 76, and a disagreement degree generating section 85 is provided with a divider that receives a disagreement degree output Δr of the error location polynomial calibration term register section 72, an estimate register section 74, and the disagreement degree generating section 85 and an error estimate term of the estimate register section 74 to calculate a new error estimate calibration term and to give it to the error location polynomial calibration term register section 72, a multiplier that receives the input of the disagreement degree output Δr from the disagreement degree generating section 85 and an input of a quotient from the error location polynomial calibration term register section 72 and multiplies them, and an adder that adds the output of the estimate register section 74 to the output of the multiplier and gives the sum to the estimated register section 74.

Description

DETAILED DESCRIPTION OF THE INVENTION TECHNICAL FIELD OF THE INVENTION

[0001] The present invention relates to a Reed-Solomon decoder,
In particular, the (Berlekamp-Massey) algorithm and the Euclidean (E
uclid) Modified Vari Kemp with an algorithm
The present invention relates to a Reed-Solomon decoding method using a Messi algorithm and its decoder.

[0002]

2. Description of the Related Art FIG. 1 shows an ATSC (American Television S).
FIG. 2 is a block diagram showing a configuration of an encoder that constitutes a forward error corrector at a transmission stage of 8VSB (Vestigial Side Band) standard. In the transmission stage having such a configuration, data is first randomized by the randomizer 11.
Then, after the Reed-Solomon encoder 12 performs Reed-Solomon encoding, an interleaver 13 interleaves for error correction, and a lattice encoder 14 lattice-encodes and transmits.

FIG. 2 is a block diagram showing a configuration of a decoder constituting a forward error corrector at a receiving stage of the ATSC 8VSB standard. In the receiving stage having such a configuration, decoding is performed by the lattice decoder (Viterbi decoder) 21 in the reverse order of transmission,
De-interleaving is performed by a de-interleaver 22. The signal output from the deinterleaver 22 is decoded by a Reed-Solomon decoder 23 and then de-randomized by a de-randomizer 24 to recover the original signal of the transmission stage.

The general decoding process of the Reed-Solomon decoder 23 is as follows. (1) 2t syndromes from received sequence r (x) signal
(syndrome). (2) An error locator polynomial (e
rror locater polynomial) and the error estimation polynomial (error eva
luator polynomial). (3) Calculate the root of the error locator polynomial to find the location of the error. (4) An error value is calculated from the error locator polynomial and the error estimation polynomial. (5) Correct an error generated in the reception sequence.

FIG. 3 is a block diagram showing the configuration of the Reed-Solomon decoder. The Reed-Solomon decoder is roughly divided into a decoding unit 31 and a delay element 32. The decoding unit 31 receives the received signal r input from the syndrome calculation unit 31a.
By substituting the root ak of the generator polynomial into (x), 2t syndrome elements S _k are obtained, and an error locator polynomial and an error estimator polynomial calculator 31b calculate an error locator polynomial and an error estimator polynomial from the calculated syndromes, respectively. . The error position and error value calculator 31c calculates an error position and an error value from the error position polynomial and the error estimation polynomial calculated by the error position polynomial and error estimation polynomial calculator 31b. The error value calculated in this way is exclusive-ORed with the received signal delayed by the time for the error value calculation of the decoding unit 31 by the delay element 32 in the exclusive OR element 33. Correct the error in the received data.

In the decoding process shown in FIG. 4, a part for calculating an error locator polynomial (an error locator polynomial and an error estimation polynomial calculator) determines the performance of the Reed-Solomon decoder. FIG. 4 is a detailed diagram of a conventional error locator polynomial calculator for calculating an error locator polynomial. As shown in FIG. 4, the timing and control signal generator 41 outputs a discrepancy (discrete)
Pancy) clock signal of each register in response to the inconsistency delta _r from generator 44 and the switch control signal Contorl
-0, Control-1 and Control-2 are generated. That is, the switch control signal Control-
0 controls the first switch S1 and the switch control signal Co
control-1 controls the second and third switches S2 and S3, and the switch control signal Control-2 controls the fourth switch S4.

The error locator polynomial calibration term register 42 receives the signal output from the divider 46 through the first switch S1, and generates an error locator calibration term from the error locator polynomial. The error position register unit 43
The error locator polynomial is generated by the signal output from the adder 47. Further, the syndrome register unit 45 receives the feedback of its own output signal and generates a syndrome term.

On the other hand, the first multiplier 49 inputs the error position term output from the error position register section 43 and the syndrome term output from the syndrome register section 45 via a fourth switch S4, and The signal output from the first multiplier 49 is
It is input to the first adder 50. The first adder 50 includes:
The output of the first multiplier 49 is added to the inconsistency from the inconsistency generator 44, and the result is input to the inconsistency generator 44.

The inconsistency generating section 44 is provided with the first adder 5
Generating a dissimilarity delta _r in accordance with a signal outputted from 0. Thus the inconsistency delta _r is generated, the divider 4
6, the timing and control signal generator 41 and the third switch are input to the first adder 50, and the divider 46 outputs the mismatch output Δ _r of the mismatch generator 44.
And an error position term output from the error position register 43, and calculates a new error position calibration term. The calculated error position term is stored in the error position polynomial calibration term register section via a first switch S1. Input to 42 to generate an error location calibration term. Also, the second multiplier 48
Is the degree of mismatch Δ output from the degree-of-mismatch generator 44
_r and the error position calibration term output from the error position calibration term register 42 are received via the second switch S2, and are multiplied and output.

The signal output in this manner is supplied to the second adder 4
7 and the second adder 47 is connected to the second multiplier 4
8 and the output of the error position register 43, and the result is input to the error position register 43. On the other hand, the first to fourth switches S1 to S4 perform a switching operation according to the switch control signals Control-0 to Control-2 output from the timing and control signal generator 41.

That is, the first switch S1 receives the output of the error locator polynomial calibration term register 42 as an input, or inputs the output signal of the divider 46 to the error locator polynomial calibration term register 42. The second switch S2 is
The output of the error locator polynomial calibration term register 42 is input to a second multiplier 48. The third switch S3 inputs the output of the inconsistency generating section 44 to the first adder 50. The fourth switch S4 inputs the output of the syndrome register unit 45 to the first multiplier 49.

Generally, for calculating the error locator polynomial, Be
The rlekamp-Massey algorithm and the Euclid algorithm are the most used, and are used to calculate the error locator polynomial.
A detailed description of the Berlekamp-Massey algorithm and the Euclid algorithm follows.

FIG. 5 shows the Berlekamp-Massey algorithm (B
FIG. 6 is a sequence diagram for calculating an error locator polynomial using MA).
Step S41 defines initial conditions of the Berlekamp-Massey algorithm BMA. That is, the first term coefficient σ of the error estimation polynomial
The initial condition of the calibration term to be added when updating the error locator polynomial is given by setting ₀ = 1 and the calibration term coefficient b ₀ = 1, and r = 0
, An initial value of the number of iterations r is determined, and an initial condition of a control signal for determining a calculation method is set as the order L = 0.

In step S42, the number of repetitions r is increased by one. At this time, if the increased number of repetitions r is 2t, in step S50, σ (x) is set as a coefficient of the error locator polynomial, and the Be
Complete the rlekamp-Massey algorithm (BMA). If the number of repetitions r is not 2t, the degree of mismatch Δr is calculated using the syndrome Sr and the coefficient σ _j of the error locator polynomial in step S43 after step S42, and the addition operation here is the addition operation in galois filed. And exclusive OR.

Next, at step S44, the degree of discrepancy (discrepanc
y: Δr) is 0, and if 0, step S
Proceeding to 48, the order of the calibration term b (x) is increased by one order. If the degree of discrepancy is not 0 in step S44, the process proceeds to step S45, where the coefficient σ _x of the error locator polynomial is σ _x −Δ
Substitute _r xb (x) to calculate a new error locator polynomial. Thereafter, in step S46, the order of the new error locator polynomial (d
egree) value L is calculated and twice (2L) of the next numerical value L is r-1.
Determine if it is greater than.

If 2L is smaller than r−1,
In step S47, the coefficient b of the calibration _r ^-1σ (x)
Substituting and substituting r−L for L, the calibration term of the error locator polynomial
And the order is updated. In contrast, the 2L is r-1
If it is larger, the process proceeds to step S48 and the error locator polynomial
Is increased by the first order. New error in step S49
Calculation of position polynomial, calibration term of error locator polynomial, control signal
After that, the value of r is increased by 1 and the value becomes 2t.
To determine At this time, if r ≠ 2t, step S4
2 and if r = 2t, σ (x) in step S50
Is calculated as the coefficient of the error locator polynomial.
finish.

When designing a Reed-Solomon decoder using such a BMA structure as it is, an error estimation polynomial is separately calculated after all the calculation of the error locator polynomial is completed. Therefore, when designing a Reed-Solomon decoder using the BMA structure, it is impossible to separately calculate an error estimation polynomial unless all the error locator polynomials are calculated, and thus the overall latency is increased. There is.
What has improved such disadvantages is an error locator polynomial and an error estimation polynomial calculation method using the Euclidean algorithm.

FIG. 6 shows an example using the Euclidean algorithm.
Order for calculating error locator polynomial and error estimation polynomial simultaneously
FIG. First, in step S51, the initial condition of the error locator polynomial is
Σ ₀(X) = 1, b_-1Set (x) = 0 and repeat
Initialize the number r = 1 and set the initial condition of the error estimation polynomial to z
_-1(X) = x2t, z₀(X) = s (x)
(Where s (x) is an error polynomial).

Next, in step S52, the error estimation polynomial (Z
_r−2 (x) / Z _r−1 (x))
t) Calculate a _r (x). In step S53, a new error location polynomial and an error estimation polynomial are calculated using the previous error location polynomial, error estimation polynomial, and quotient obtained in the previous step. That is, the calibration section _{σ r (x) σ r-} 2 (x) -a r
(X) σ _r-1 (x) is substituted, and z _r-2 is substituted for z _r (x).
_(X) substituting _{-a r (x) z r -1} (x). Stage S
At 54, it is determined whether the order of the error estimation polynomial z _r (x) is smaller than t.

If it is determined that the order of z _r (x) is not smaller than t, a new error locator polynomial is calculated in step S55, the number of repetitions is increased by 1, and the process from step S52 is repeated. If the order of the determination result z _r (x) is smaller than t, the final value of σ (x) becomes an error locator polynomial after repeatedly calculating t times in step S56.
The value of (x) becomes the error estimation polynomial. The highest order of the error locator polynomial calculated using the Euclidean algorithm is t, and the highest order of the error estimation polynomial is t-1.

[0021]

[Problems to be solved by the invention] By the way, Berlekamp-
When designing the Reed-Solomon decoder using the Massey algorithm, the size of the hardware is smaller than that of the Euclidean algorithm, but since the error locator polynomial is calculated, the error estimation polynomial can be obtained.
There is a disadvantage that the entire decoding delay time is longer than that of the Euclidean algorithm. On the other hand, the Euclidean algorithm has a short decoding delay time overall, but has a disadvantage that the size of hardware is large.

The present invention is intended to solve various problems arising from a conventional Reed-Solomon decoder using the Berlekamp-Massey algorithm and a Reed-Solomon decoder using the Euclidean algorithm, respectively. It is an object of the present invention to provide a Reed-Solomon decoding method using a modified Berlekamp-Massey algorithm, which is a compromise between the -Massey algorithm and the Euclidean algorithm, and its decoder.

Another object of the present invention is to provide a Berlekamp-Ma
Modified Berlekamp-Ma modified ssey algorithm to reduce decoding time delay and hardware size
An object of the present invention is to provide a Reed-Solomon decoding method using the ssey algorithm and a decoder therefor.

[0024]

[MEANS FOR SOLVING THE PROBLEMS] To achieve the above object
The Reed-Solomon decoder according to the present invention
The input signal of each register is determined according to the degree of coincidence and the input clock.
The timing and timing of generating the switch control signal
And a control signal generator and a calibration term for the error locator polynomial.
Error locator polynomial calibration term register and error locator polynomial
Error location register that generates
A syndrome register section, and the syndrome register.
Output of the syndrome term generated from the register and the error
The output of the error position term generated from the position register
And the degree of mismatch Δ_rWith a mismatch generator that generates
Calibration term of error estimation polynomial
The error estimation polynomial calibration term register that calculates
An error estimation register for generating a constant polynomial,
Output from the degree generator_rAnd the error estimation register section
Calculates a new error estimation calibration term with the input error estimation term
Input to the error estimation polynomial calibration term register section.
And a discrepancy degree Δ from the discrepancy degree generation unit. _rInput
Receiving a quotient from the error locator polynomial calibration term register.
A multiplier for receiving and multiplying the force, and
The output of the error estimation register section is added and the resulting value is
And an adder for inputting to the error estimation register section.
Is done.

A modified Berlekamp-Massey according to the invention
The Reed-Solomon decoding method using the algorithm has a calibration term b ₀ = 1, an error location term σ ₀ = 1, the number of iterations r = 0,
Initialize by setting the order L = 0, the quotient a ₀ = 1, the error estimation term z ₀ = 0, and the degree of mismatch Δ ₁ = s ₀ (s ₀ : the initial value of the syndrome term), and increase the repetition rate r by one. a first step, multiplied by the error position claim sigma _j and the syndrome term S _j, a second step of calculating the inconsistency delta _{r + 1} by adding the multiplication result and the degree of mismatch, inconsistency degree delta _r is "0" If there are a a judge inconsistency delta _r is not "0", the calibration section b (x)
And multiplied by the inconsistency delta _r, the multiplication result (-) to the error position section by adding the error position claim sigma _x as the code sigma _x - [delta _r
xb (x), the error estimation polynomial calibration term a _r (x)
And the degree of inconsistency Δr, and the result of the multiplication is added as a (−) code to the error estimation term z _x to add z _x −Δ _r to the error estimation term.
a third step of substituting a _r (x), a fourth step of calculating the degree L of the new error locator polynomial and determining whether L is greater than r−1, and 2L is determined by the fourth step. r-1
If not larger, the degree of mismatch Δ _r is divided by the error position term σ (x), and the resulting value Δ _r ⁻¹ σ (x) is substituted for the calibration term b, and the error estimation term z (x) does not match. dividing the degree delta _r, the result value of delta _r ^{-1 xz} (x) is substituted into the error estimation calibration section a, the quotient of the error locator polynomial by substituting r-L L, and calibration terms and error estimation section And if the 2L is greater than r-1 in the determination of the fourth step, the calibration term b of the error locator polynomial and the calibration term a of the error estimation polynomial are increased by one order, and the new error locator polynomial is increased. , Calibration terms,
After the calculation of the control signal is completed, the value of the repetition factor r is increased by 1 to determine whether the value has reached 2t, and if the repetition factor r is not 2t in the determination of the sixth step, First
The process returns to the step of increasing the degree of repetition r of the step, and if r = 2t, a seventh step in which σ (x) is an error position term and z (x) is an error estimation term.

[0026]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the operation and operation of the present invention will be described in detail with reference to preferred embodiments of the present invention based on such technical ideas. FIG. 8 shows the modified Be according to the present invention.
A block diagram showing the configuration of a Reed-Solomon decoder using the rlekamp-Massey algorithm is shown. The timing and control signal generator 71 receives the input clock signal as shown in FIG. 9A and the inconsistency Δr from the inconsistency generator 85.
And the switch control signals Control-0, Control-1, and Control
rl-2 is generated.

That is, the switch control signal Contro
1-0 calculates a degree value L of the new error locator polynomial and determines whether the order value 2L is greater than r-1 and determines 2
If L> r-1, a high signal ("1") is output as a control signal for the first and second switches S1 and S2 for t + 2 clocks in the clock as shown in FIG.
If L <r-1, a low signal ("0") is output as a control signal for the first and second switches S1 and S2 for t + 2 clocks.

The switch control signal Control-
1 is a high signal ("1") only during the t + 2 clock.
And controls the third to fifth switches S3 to S5, and the switch control signal Control-2 is output as a high signal ("1") only at the time point t + 1 of the input clock as shown in FIG. To control the sixth switch S6. The error estimation polynomial calibration term register 72 receives the signal output from the first divider 77 through the first switch S1, and calculates an error estimation calibration term from the error estimation polynomial.

The error locator polynomial calibration term register 73 receives the signal output from the second divider 78 through the second switch S2, and generates an error locator calibration term from the error locator polynomial. The error position register unit 75 generates an error position polynomial from the signal from the second adder 80, and the error estimation register unit 74 generates an error estimation polynomial from the signal from the first adder 79. Further, the syndrome register unit 76 receives the feedback of its own output signal and generates a syndrome term.

On the other hand, the third multiplier 84 receives an error position term output from the error position register section 75 and a syndrome term output from the syndrome register section 76 through a sixth switch S6, and receives two terms. The output signal is multiplied, and the signal output from the third multiplier 84 is input to the third adder 81. The third adder 81 adds the output of the third multiplier 84 and the degree of inconsistency from the inconsistency generator 85, and inputs the result to the inconsistency generator 85.

The non-coincidence degree generating section 85 is provided with the third adder 8
Generating a dissimilarity delta _r in accordance with a signal output from the 1. Thus the inconsistency delta _r generated first and second
The signals are input to the dividers 77 and 78 and the first and second multipliers 82 and 83, respectively. The first divider 77 calculates a new error estimate calibration section receives the input of the error estimation section that is output from the disagreement of the output delta _r error estimating register unit 74 of the inconsistency generating unit 85, the calculation The error estimation term obtained is input to the error estimation polynomial calibration term register section 72 via the first switch S1 so that the error estimation calibration term is generated.

Further, the second divider 78 generates the degree of inconsistency.
Discrepancy Δ output from component 85 _rAnd error position cash register
Receiving the error position term output from the
New calibration term, and the error location term thus calculated
Is the error locator polynomial calibration term through the second switch S2.
Transmitted to the register 73 to create a new error calibration term
To be

The first multiplier 82 generates the mismatch degree
Discrepancy Δ output from component 85 _rInput,
The error estimation polynomial calibration term output from the register unit 72
Input of the error estimation calibration term through the third switch S3.
And multiply and output. The signal thus output is
The signal is input to a first adder 79, and the first adder 79
The output of the 1 multiplier 82 and the output of the error estimation register 74
And adds the result to the error estimation register 74.
Power.

The second multiplier 83 is provided with the inconsistency degree generating section 8
5 receives the inconsistency delta _r outputted from said receiving an input of the error locator polynomial calibration section that is output from the error position polynomial calibration section register section 73 through the fourth switch S4, and outputs the multiplication. The signal output from the second multiplier 83 is input to a second adder 80 and the second adder 8
0 adds the output of the second multiplier 83 and the output of the error position register unit 75 and inputs the result to the error position register unit 75.

On the other hand, the first to sixth switches S1 to S6 are
Switch control signals Control-0 to Control output from the timing and control signal generator 71
The switching operation is performed according to -2. That is, the first switch S1 is connected to the error estimation polynomial calibration term register 72.
Is input again, or the output signal of the first divider 77 is input to the error estimation polynomial calibration term register 72. The second switch S2 receives the output of the error locator polynomial calibration term register 73 as an input or inputs the output signal of the second divider 78 to the error locator polynomial calibration term register 73. The third switch S3 inputs the output of the error estimation polynomial calibration term register unit 72 to the first multiplier 82. The fourth switch S4 inputs the output of the error locator polynomial calibration term register unit 73 to the second multiplier 83. The fifth switch S5 is connected to the inconsistency generation unit 85
Is input to the third adder 81. The sixth switch S6 outputs the output of the syndrome register 76 to a third
Input to the multiplier 84.

FIG. 7 shows a modified Berlekamp- according to the invention.
FIG. 4 is a flowchart illustrating a Massey algorithm. Step S61
In the initialization process, the calibration term b ₀ = 1, the error location term σ ₀ = 1, the number of iterations r = 0, the order L = 0, the error estimation calibration term a ₀ = 1, and the error estimation term z _{0 are as} follows. = 0, degree of mismatch Δ ₁ = s
₀ (s ₀ : initial value of the syndrome term) is set. In step 62, the repetition factor r is increased by one. In step S63, the error position term σ _j output from the error position register unit 74 and the syndrome S _j output from the syndrome register unit 76 are multiplied by a third multiplier 84 via a sixth switch S6. Output of Multiplier 84 and Dissimilarity Generation Unit 8
The output of 5 are added by the third adder 81 via the fifth switch S5 to calculate the inconsistency delta _r by the following formula.

[0037]

(Equation 1)

In step S64, the degree of mismatch Δ_rIs "0"
It is determined whether or not there is. Discrepancy Δ _rIs not "0"
In step S65, the second multiplication is performed through the fourth switch S4.
The error position calibration term b (x) is applied to the detector 83 and the degree of mismatch Δ
_rAnd the result of the multiplication as the (−) sign
The error position term σ in the arithmetic unit 80_xAnd error position register
Σ_x−Δ_r× b (x), and the third switch
The error estimation calibration term a is supplied to the first multiplier 82 via the switch S3.
_r(X) is applied to the discrepancy Δ_rAnd multiply by
The first adder 79 sets the error estimation term z as the (-) code as the result.
_xIs added to the error estimation register 74._x−Δ_ra
_rSubstitute (x).

In step S66, the order value L of the new error locator polynomial is calculated to determine whether the order value 2L is greater than r-1. As a result of this determination, if the order value 2L is smaller than r-1, the error position term σ is stored in the second divider 78 in step S67.
(X) and the degree of mismatch Δ _r in the error position term σ (x)
Dividing a resulting delta _r ^-1 sigma (x) is substituted into the calibration section b, the inconsistency in the error estimating section z (x) by applying an error estimation section z (x) in the first divider 77 delta _r and the resulting Δ
Substituting _r ⁻¹ xz (x) into the error estimation calibration term a,
-L is substituted to update the order of the error locator polynomial, the calibration term and the error estimation term.

If it is determined that the order value 2L is larger than r-1, the calibration term b and the error estimation term a of the error locator polynomial are increased by one order in step S68. After calculating the new error locator polynomial, calibration term, and control signal in step S69, r
Is incremented by 1 and it is determined whether the value has reached 2t. If r ≠ 2t, the process returns to step S42, where r = 2t
Then, in step S70, σ (x) is set as an error position term, and z
Let (x) be the error estimation term.

[0041]

As described above, according to the present invention,
The modified Berlekamp Massey algorithm has the advantage that the hardware of the Reed-Solomon decoder can be easily calibrated. Also, a modified Berlekamp-Mass
With the ey algorithm, the error locator polynomial and the error estimator polynomial can be calculated simultaneously, which has the effect of reducing the overall decoding delay time.

[Brief description of the drawings]

Fig. 1 ATSC (American Television System Comm
FIG. 4 is a block diagram showing a configuration of an encoder that constitutes a forward error corrector at a transmission stage of 8VSB (Vestigial Side Band) standard.

FIG. 2 is a block diagram illustrating a configuration of a decoder that constitutes a forward error corrector at a receiving stage of the ATSC 8VSB standard.

FIG. 3 is a block diagram illustrating a configuration of a Reed-Solomon decoder.

FIG. 4 is a detailed diagram of an error locator polynomial calculator (31b) of FIG. 3;

FIG. 5: Varikemp-Messi of Reed-Solomon decoder
FIG. 4 is a sequence diagram for obtaining an error locator polynomial using a (Berlekamp-Massey) algorithm (BMA).

FIG. 6 is a flowchart illustrating an algorithm for simultaneously calculating an error locator polynomial and an error estimation polynomial using the Euclid algorithm.

FIG. 7 is a flowchart illustrating a modified VariKemp-Messi algorithm according to the present invention.

FIG. 8 is a block diagram showing a configuration of a Reed-Solomon decoder using a modified Variken-Messi algorithm according to the present invention.

FIG. 9 is a timing chart of the control signal generated from the timing and control signal generator of FIG. 8;

[Explanation of symbols]

11 randomizer, 12 Reed-Solomon encoder, 13
Interleaver, 14 lattice encoder, 21 lattice decoder, 22 inverse interleaver, 23 Reed-Solomon decoder, 24 inverse randomizer, 13 interleaver, 14
Lattice encoder, 31 decoding unit, 31a syndrome calculation unit, 31b error location polynomial and error estimation polynomial calculation unit, 31c error location and error value calculation unit, 32 delay element, 33 exclusive OR element, 41 timing and control signal generation Unit, 42 error position polynomial calibration term register unit, 43 error position register unit, 44 inconsistency generation unit, 45 syndrome register unit, 46 divider, 4
7 Second adder, 48 Second multiplier, 49 First multiplier, 50 First adder, S1 first switch, S2 second switch, S3 third switch, S4 fourth switch, 71 Timing and control signal generation Section, 72 error estimation polynomial calibration term register section, 73 error location polynomial calibration term register section, 74 error estimation register section, 75 error location register section, 76 syndrome register section, 7
7 First divider, 78 Second divider, 79 First adder, 80 Second adder, 81 Third adder, 82 First
Multiplier, 83 Second multiplier, 84 Third multiplier, 85
The dissimilarity generation unit, S5 fifth switch, and S6 sixth switch.

Claims (7)

[Claims] 1. A timing and control signal generator for generating a control signal of a switch for determining an input signal of each register according to the generated degree of mismatch and an input clock, and an error position for generating a calibration term of an error position polynomial A polynomial calibration term register section, an error location register section for generating an error location polynomial, a syndrome register section for generating a syndrome term, an output of the syndrome term generated from the syndrome register section, and an error generated from the error location register section. in Reed-Solomon decoder having a inconsistency generator for generating a dissimilarity delta _r the output of the position terms as input, and the error estimation polynomial calibration term register unit for calculating the calibration section of the error estimation polynomial, generating an error estimation polynomial an error estimator register unit which, said error estimation les an inconsistency output delta _r of the inconsistency generator A divider to be input to the error-estimating polynomial calibration section register unit calculates a new error estimate calibrate claim error estimation section register unit as input, inconsistency delta _r and the error position polynomial calibration from the inconsistency generator A multiplier that multiplies the quotient from the term register unit as an input, and an adder that adds the output of the error estimation register unit to the output of the multiplier and inputs the result value to the error estimation register unit. A Reed-Solomon decoder using a modified Berlekamp Massey algorithm, wherein the decoder is configured. 2. The error estimation calibration term a _r (x) in the multiplier.
And multiplied by the inconsistency delta _r, the multiplication result (-) as a code, the substituting _{z x -Δ r a r (x} ) to the error estimation section by adding the error estimated claim z _x by an adder The Reed-Solomon decoder using the modified Berlekamp Massey algorithm according to claim 1. 3. A degree L of a new error locator polynomial is calculated to determine whether a degree 2L is larger than r-1. If not, an error estimation z (x) is applied to a divider to estimate an error. The degree of mismatch Δ _r is divided by the term z (x), and as a result Δ _r
^-1 z (x) is substituted for the error estimation calibration term a, and L is r−L
2. The Reed-Solomon decoder using the modified Berlekamp Massey algorithm according to claim 1, wherein the order of the error estimation term is updated by substituting the following. 4. A first step of setting initial conditions for generating an error locator polynomial and an error estimation polynomial in a Reed-Solomon decoding method, and after setting the initial conditions, an error locator term σ _j and a syndrome term S _j. multiplied by a second step of calculating the degree of mismatch by adding the results and inconsistency of the multiplication, the calculated discrepancy degree delta _r is "0" is either a judge inconsistency delta _r is "0 If not, a third step of calculating a new error locator polynomial and an error estimation polynomial;
A fourth step of calculating the degree L of the new error locator polynomial and determining whether L is greater than r-1; if the result of the fourth step is that 2L is not greater than r-1, the error locator polynomial And the fifth step of updating the quotient of the calibration term and the error estimation term, and if the result of the fourth step 2L is greater than r-1, the error location calibration term b and the error estimation calibration term a of the error locator polynomial Is increased by one order, and after the calculation of the quotient and the calibration term of the new error locator polynomial and the control signal is completed, the repetition rate r
A step of incrementing the value of by 1 and determining whether the value has reached 2t, and, if the repetition rate r is not 2t in the determination of the sixth step, increasing the repetition rate r of the first step And if r = 2t, σ (x) is taken as the error position term,
a modified Berlekamp Massey algorithm using a modified Berlekamp Massey algorithm. 5. The first step of setting the initial condition includes a calibration term b ₀ = 1, an error location term σ ₀ = 1, a number of iterations r = 0,
Order L = 0, error estimation calibration term a ₀ = 1, error estimation term z ₀
5. The initial condition is set by setting = 0 and the degree of mismatch Δ ₁ = s ₀ (s ₀ : initial value of a syndrome term), and initializing by increasing the degree of repetition r by one. Reed-Solomon decoding method using modified Berlekamp Massey algorithm. 6. The error correction step b
(X) is multiplied by the degree of inconsistency, and the multiplication result is represented by
Error location term σ_xAnd add σ to the error location term _x−Δ
_rxb (x), the error estimation calibration term a_r(X) and not
Matching degree Δ_rAnd the result of the multiplication as a (-) sign
Error location term z_xAnd add z to the error estimation term_x−Δ_ra
_rSubstitute (x) for new error locator polynomial and error estimation
5. A modification according to claim 4, wherein a polynomial is generated.
Lead source using the modified Berlekamp Massey algorithm
Romon decoding method. 7. In the fifth step, the degree of mismatch Δ _r is divided by the error location term σ (x), and as a result, Δ _r ⁻¹ σ (x) is substituted for the error location calibration term b, and the error estimation term is calculated. dividing the inconsistency delta _r in z (x), the result delta _r ^{-1 xz} (x) is substituted into the error estimation calibration section a, the quotient and calibration terms of the error locator polynomial by substituting r-L to L 5. The modified Berlekamp Ma according to claim 4, wherein the order of the error estimation term is updated.
A Reed-Solomon decoding method using the ssey algorithm.