During the transmission of digital signals, errors (error codes) of the transmitted digital signals may occur due to interference or channel characteristics deterioration, etc. Therefore, error correction is an essential process to improve the quality of digital transmission.
So how does error correction work?After the modulo/digital transformation of the analog signal, the additional data (such as the parity bit) is added to the data stream. At the receiving end, the wrong data word is found through the parity bit (that is, the wrong data word is identified by the identification of the additional data) and is corrected.There are three ways to correct errors: noise, keeping the word in front, and linear interpolation.
When an error occurs and is identified, the circuit concerned will terminate the transmission of the circuit at the point of the error, that is, to solve the noise.Noise reduction only works in the error correction process and is usually used in the case of continuous errors.
Digital processing circuits typically use a large number of memory circuits to process digital signals for a brief period of continuous memory, so that when a "suspicious" data word is found, it can be replaced by the nearest data.Replacing the incorrect data word with the preceding data word results in fewer errors (due to correlation) than the data without errors, so this error correction method is reasonable.So keeping the front edge is an acceptable way to correct errors.
Linear interpolation can further improve the above error correction method.The so-called linear interpolation is to take the average value of the previous data word and the next data word of the error word, and replace the error word with this value, obviously this method can get more accurate error correction.