JP2636432B2 - Error correction method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error correction method used in digital mobile radio, moving image compression transmission, and the like.

2. Description of the Related Art In general, the minimum value of the Hamming distance between codewords of an error correction code having a code length n (= k + m) obtained by adding m check bits to k information bits is referred to as a minimum distance of the error correction code. , This minimum distance is 2t + 1 (t is a positive integer)
In this case, an error of t bits or less can be correctly corrected.

Here, there are a total of 2 ^k types of codewords, but when a sphere of radius t centered on each codeword fills the n-dimensional space without any gap, Holds, this error correction code is called a complete code, Holds, this error correction code is called an incomplete code.

In the case of a non-perfect code, even if an error of t + 1 bits or more occurs in a certain transmission word, the reception word does not always fall within the Hamming distance t from another transmission word. That is, the non-perfect code can not only correct an error of t bits or less, but also correct an error of t + 1 bits or more without erroneous correction, that is, without performing incorrect correction.
In some cases, it can be detected that there is an error of one or more bits.

FIG. 3 is a flowchart showing a conventional error correction method for a t-double error correction incomplete code.

In FIG. 3, the received transmission signal is reproduced into a binary signal by a threshold value for determining whether the received bit is “0” or “1”. In the following step 24, an error of t + 1 times or more is detected.

Here, if an error of t times or more occurs and the received word is within a Hamming distance t from another transmitted word, the error has been erroneously corrected in step 23, and no error is detected in step 24. The process ends (step 26).

On the other hand, if the received word does not fall within the Hamming distance t from another transmitted word, it can be detected in step 24 that an error of t + 1 times or more has occurred. A flag indicating that an error is included is set, and the error is terminated (step 26).

Therefore, in the above-mentioned conventional example, an error of t bits or less can be correctly corrected, and when an error of t + 1 bits or more occurs, when the error does not fall within the Hamming distance t from another transmission word, t + 1 bits or more Can be detected.

SUMMARY OF THE INVENTION However, in the above-described conventional error correction method, when an error of t + 1 bits or more that exceeds the maximum number of error correction bits t occurs, the occurrence can be detected.
There is a problem that the error cannot be corrected.

The present invention has been made in view of the above-described conventional problems, and when an error of t + 1 bits or more that exceeds the maximum error correction bit number t occurs, even when a received word is not within a Hamming distance t from another transmitted word that is not a correct transmitted word. It is another object of the present invention to provide an error correction method capable of correctly correcting an error.

Means for Solving the Problems In order to achieve the above object, the present invention receives an incomplete error correction coded bit sequence having a code length of n bits and a maximum number of error correction bits of t bits, and generates the bit sequence from a received word. The error of t bits or less is corrected by the generated syndrome, and when an error of t + 1 bits or more is detected in the error correction, t + 1 (i = 1, 2,...) Is set with i = 1 as an initial value.
I) A bit error pattern is added to a received word to determine whether or not the syndrome is “0” for all patterns. If there is only one error pattern, the error pattern of the received word is corrected by t + i bits. It is like that.

The present invention also receives an incomplete error correction coded bit string having a code length of n bits and a maximum number of error correction bits of t bits, and sets a flag to the received bit if the bit has more than a predetermined noise. Then, an error of t bits or less is corrected by a syndrome generated from the received word, and when an error of t + 1 bits or more is detected in this error correction, an error pattern of t + 1 bits is added to the received word and the syndrome becomes "0". Is determined, and if the number of error patterns is one, the error pattern of the received word is corrected by t + 1 bits. The error pattern of the received word having the largest number of bits is t + 1-bit error-corrected.

According to the present invention, an error pattern is added when an error of t + 1 bits or more exceeding the maximum error correction bit number t occurs, so that a received word is within a Hamming distance t from another transmitted word that is not a correct transmitted word. Even when it does not become
Errors can be corrected correctly.

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a flowchart showing one embodiment of the error correction method according to the present invention.

The transmission signal received in step 1 shown in FIG. 1 is first reproduced in step 2 into a binary signal by a threshold value for determining whether the received bit is “0” or “1”. Is correctly corrected, and it is then detected in step 4 whether or not there is an error of t + 1 or more.

Here, if an error of t times or more occurs and the received word is within a Hamming distance t from another transmitted word, the error has been erroneously corrected in step 3 and the error is detected in step 4. And end the process (step 1
1).

On the other hand, if the received word does not fall within the Hamming distance t from another transmitted word, the process proceeds to step 5 where i (i is a positive integer) is initialized (i = 1). Add a bit error pattern to the received word,
The number N of error patterns in which the syndrome becomes “0” is obtained.

Next, in step 7, if N ≧ 2, it is not possible to determine which error pattern is the t + i-bit error pattern generated in the received word, so the process ends (step 11). If = 1, the process proceeds to step 8, the pattern is determined as an error pattern, and t +
Perform i-level error correction. On the other hand, when N = 0,
Since the error is not t + i bits, the process branches to step 9.

In step 9, it is determined whether or not a predetermined positive integer I is equal to i based on the weight distribution of the code word, the time allowable in the correction processing, and the like.
Proceeding to 10, increment i by one, return to step 6, and repeat the above operation. In step 9, i = I
If so, the process ends (step 11).

Therefore, according to the above embodiment, an error pattern of t + i bits is added to a received word, so that an error of t + 1 or more can be corrected in spite of a t-multiple error correction code, and the error correction capability is improved. be able to.

 FIG. 2 shows a second embodiment of the present invention.

When a transmission signal is received in step 12, first, in step 13, it is determined whether or not the received bit has more noise than a specified value. still. This determination is made based on whether or not the received bit is close to a threshold for determining “0” or “1”.

In step 15, a binary signal is reproduced by the threshold value. In step 16, errors of t or less are correctly corrected. Then, in step 17, it is detected whether there is an error of t + 1 or more.

When an error of t + 1 times or more is detected, the process proceeds to step 18, where an error pattern of t + 1 bits is added to the received word,
The number N of error patterns in which the syndrome becomes “0” is obtained.

If N ≧ 2, the process proceeds to step 21 where the number of bits with the flag set is counted for each t + 1 bit error pattern, and the pattern with the largest number of bits with the flag set is determined to be an error pattern. , T + 1 double error correction.

If N = 1, the process proceeds to step 20, the pattern is determined to be an error pattern, and t + 1 double error correction is performed.
If N = 0, the process ends (step 22).

Therefore, in the above-described embodiment, a flag is set in advance to a reception bit having noise equal to or greater than a prescribed value, and an error pattern of t + 1 bits is added to the reception word. The error can be corrected, and the error correction capability can be improved.

As described above, according to the present invention, when an error of t + 1 bits or more that exceeds the maximum number of error correction bits t occurs,
Since the error pattern is added, even when the received word does not fall within the Hamming distance t from another transmitted word that is not a correct transmitted word, the error can be correctly corrected.

[Brief description of the drawings]

FIG. 1 is a flowchart showing an embodiment of the error correction method according to the present invention, FIG. 2 is a flowchart showing a second embodiment of the error correction method according to the present invention, and FIG. 9 is a flowchart illustrating a correction method.

Claims (3)

(57) [Claims] An incomplete error correction coded bit string having a code length of n bits and a maximum error correction bit number of t bits is received, and an error of t bits or less is corrected by a syndrome generated from a received word. , T + 1 in the error correction
When an error of more than bits is detected, a syndrome for a pattern in which an error pattern of t + i (i = 1, 2,... I; I is a positive integer) bits is added to a received word with i = 1 as an initial value is obtained. It is determined whether the syndrome becomes “0”. If only one of all t + i bit patterns exists in the syndrome, the added t + i
An error correction method characterized in that a bit pattern is regarded as an actually generated error pattern and t + i-bit error correction is performed. 2. A non-perfect error correction coded bit string having a code length of n bits and a maximum number of error correction bits of t bits is received, and if the received bits have more than a predetermined level of noise, a flag is set in the bits. Then, an error of t bits or less is corrected by a syndrome generated from the received word, and when an error of t + 1 bits or more is detected in the error correction, an error pattern of t + 1 bits is added to the received word to make the syndrome " 0 is determined, and if there is one error pattern in which the syndrome is “0”, the error pattern of the received word is corrected by t + 1 bits, and the error pattern in which the syndrome is “0” is 2 In these cases,
An error correction method comprising: correcting an error pattern of a received word having the largest number of bits with a flag set by t + 1 bits. 3. The method according to claim 2, wherein whether the received bit has noise equal to or greater than a predetermined value is determined based on whether the received bit is close to a threshold value for determining whether the received bit is “0” or “1”.
Error correction method described.