Definition
Even parity (ECC) is a way to correct data errors during data transmission. It is divided into two types: odd parity and even parity.
If odd parity is used, an additional bit is added as a check bit when transmitting each byte. When the number of "1"s in the actual data is even, this check bit It is "1", otherwise the parity bit is "0", which ensures that the transmitted data meets the requirements of odd parity. When the receiver receives the data, it will detect the number of "1"s in the data according to the requirements of odd parity. If it is an odd number, it means the transmission is correct, otherwise it means the transmission is wrong.
Similarly, the process of even parity is the same as that of odd parity, except that the number of "1"s in the detection data is even, that is, the added check bit makes the number of 1 or 0 in the code The number is even.
Example
For example, the even parity code of 0100101 is 010010111.
Inference even parity: When the number of "1" in the actual data is even, the parity bit is "0", otherwise the parity bit is "1", so that it can be guaranteed The transmitted data meets the requirements of even parity. When the receiver receives the data, it will detect the number of "1"s in the data according to the requirements of even parity. If it is an even number of "1"s, it means the transmission is correct, otherwise it means the transmission is wrong.
Row and Column Check Code
The row and column check code is also called the horizontal and vertical check code or the two-dimensional parity check code, sometimes also called the matrix code. It not only performs a parity check on the symbols in the horizontal (row) direction, but also on the symbols in the vertical (column) direction. Generally, there are L×m information elements, and L+m+1 check elements are added. L+1 rows and m+1 columns form a codeword of (Lm+L+m+1, Lm) row and column check code. Table 8-2 is a codeword (L=5, M=10) of the (66, 50) row and column check code. Each row and each column performs an even check for the number of l. It can be transmitted row by row or column by column. When decoding, check the check relationship of each row and each column to determine whether there is an error.
This code may detect an even number of errors. Although the parity bit of each row cannot be used to detect the even number of error codes in the row, it is possible to detect it in the direction of the column. However, there are some even-numbered error codes that cannot be detected. For example, the four error codes that form a rectangle cannot be detected.
This two-dimensional even check code is suitable for detecting burst error codes. Because this kind of burst error codes often appear in series, followed by a longer error-free interval, there are more chances of multiple odd or even error codes in a certain row. This square matrix code is suitable for detecting such errors. Code. The aforesaid one-Vicchi check code is generally only suitable for detecting random errors.
Since the square matrix code can only detect the error codes forming the four corners of the rectangle, its error detection ability is strong. Some test measurements have shown that this code can reduce the bit error rate from one percent to one ten thousandth of the original bit error rate.
The two-dimensional parity check code can be used not only to detect errors, but also to correct some error codes. For example, when there are odd errors in only one row in the code group, the position of the error code can be determined, so as to correct it.
Distribution matrix
The number of bits P of an original sparse sequence shifted periodically is determined by the code rate required by the system and the number of sequences t used. Since the code is quasi-cyclic, its check matrix can be regarded as a combination of some cyclic sub-matrices after the column position is exchanged. Each sub-matrix can be obtained by sampling the columns of the original check matrix, and the sampling period is p. A sparse sequence can also be regarded as a combination of its p sampling sequences.
If you can get another distribution matrix by only exchanging the rows or columns of one distribution matrix, we say that the two matrices are isomorphic.
Generate
When t good sparse sequences are obtained, it is easy to construct the check matrix of sparse sequence codes. First, take t sparse sequences as t key rows of the check matrix. Secondly, by cyclically shifting the key row to the left, shifting p positions each time without repeating it, and obtaining t sub-matrices. Finally, the check matrix of the LDPC code is obtained by connecting the t sub-matrices up and down.
Verification
The domestic regulations stipulate in JJG1021-1990 "Product Quality Inspection Agency Metrology Certification Technical Assessment Specification" and some other documents: if there is no verification procedure, it should be written by the enterprise The verification method is verified. In 4.11 of the ISO9001 standard, the word "verification" also appears in many places. For example, "If the test software or comparison standard is used as a means of inspection, it should be checked before use, ... and rechecked at the specified period." Analysis of the usage of verification at home and abroad, its meaning is basically The same, it has a certain connection with verification and calibration, but there are obvious differences.
It does not have the same legality as calibration, and it has commonality with verification in terms of technical operation content. Generally, calibration can be carried out, and other related performance can also be inspected, and the eligibility is finally given. The conclusion. This term is sometimes necessary. It is recommended that a formal position should be given to the verification in the definition of related terms to unify and standardize its use.