JPS62281531A - Difference pulse coding system - Google Patents

Abstract

PURPOSE: To keep a signal error recovered at a decoder side within a prescribed period even if an error takes place in difference information on a transmission line by sending part of forecast signal in addition to the difference information from the coder side to the decoder side. CONSTITUTION:A difference signal 22 quantized by a quantization circuit 1 is coded by a coding circuit 3 and a code 24 being the difference information is written in a difference information memory 7 by a write circuit 4. The difference signal 22 is added to a forecast signal 21 at the same time to form a recovered signal 23. A decoding circuit is operated entirely opposite to the operation of the coding circuit 3 and outputs a difference signal. The difference signal is added to the forecast signal and the result is outputted at an output terminal as a recovered signal. If an error takes place in the difference information, the error of the recovered signal is kept within the time till the next forecast signal sent from the coder is used and all the succeeding signals are not in error as a conventional system.

Description

[Detailed description of the invention] 3. Detailed description of the invention [Industrial application field] The present invention relates to a differential pulse encoding method.

(Prior Art) In a conventional differential pulse encoding method (including an adaptive differential pulse encoding method), a difference value between an input signal value and a predicted signal value based on a past signal value is encoded and transmitted. On the decoding side, this difference value is added to the predicted signal value based on the past difference value to reproduce the signal.

[Problem that the invention seeks to solve]

In the conventional differential pulse encoding method described above, the information transmitted from the encoding side to the decoding side is only the difference value between the input signal value and the predicted signal value, and on the decoding side, all past differences Since the signal is reproduced based on the value, if an error occurs even once in the information on the transmission path, no matter how correctly the subsequent information is transmitted, all the signals reproduced thereafter will be erroneous.

SUMMARY OF THE INVENTION An object of the present invention is to provide a differential pulse encoding method that allows accurate transmission of information.

(Means for Solving the Problems) The differential pulse encoding method of the present invention has a means for transmitting some predicted signal values in addition to the difference information on the encoding side, and a means for transmitting some predicted signal values on the decoding side. When reproducing a signal by adding a difference value to the encoder, if a predicted signal value is transmitted from the encoding side, the difference value is added to this predicted signal value, and if the predicted signal value is not transmitted, the past value is added to the predicted signal value. The present invention is characterized by having means for adding a difference value to a predicted signal value based on the difference value.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

1 and 2 are block diagrams showing one embodiment of the present invention, and FIG. 1 is a block diagram of the encoding side.

FIG. 2 is a block diagram of the decoding side. Figures 1 and 2
As shown in the figure, in this embodiment, information is transmitted from the encoding side to the decoding side by writing and reading from memory.

First, the operation on the encoding side will be explained using FIG. A difference value between the input signal value 20 input to the input terminal 10 and the predicted signal value 21 is input to the quantization circuit 1 . The difference signal value 22 quantized by this circuit 1 is sent to an encoding circuit 3.
, and becomes the code 24 which is the difference information. Reference numeral 24 is written into the difference information memory 7 by the write circuit 4. At the same time, the difference signal value 22 is added to the predicted signal value 21 to obtain a reproduced signal value 23. In addition, the reproduction signal value 23
As will be described later, this is exactly the same as what is reproduced and output on the decoding side. Prediction circuit 2 reproduces signal value 23
The next input signal value 20 is predicted based on . The prediction circuit 2 outputs this prediction result as a prediction signal value 21 when the next input signal value 20 is input.
This is repeated each time the input signal value 20 is input.

On the other hand, the predicted signal value 21 is written into the predicted signal value memory 8 by the write circuit 5 controlled by the control circuit 6. Not all predicted signal values are written to this memory, but some predicted signal values, for example, periodically at a rate of 1 out of 100 predicted signal values.

Next, the operation on the decoding side will be explained using FIG.

The code 40 read out from the difference information memory 30 by the reading circuit 32 is input to the decoding circuit 33.

The decoding circuit 33 performs an operation completely opposite to that of the encoding circuit 3 in FIG. 1, and outputs a difference signal value 41. The difference signal value 41 is added to the predicted signal value 45 and outputted to the output terminal 50 as a reproduced signal value 42. The selection circuit 36 is controlled by the control circuit 35, selects either the predicted signal value 43 or the predicted signal value 44, and outputs it as the predicted signal value 45. If the predicted signal value necessary to obtain the reproduced signal value 42 is stored in the predicted signal value memory 31, the control circuit 35 controls the readout circuit 34 to read the predicted signal value from the memory.

At the same time, the selection circuit 36 is controlled to select this predicted signal value 44. If the required predicted signal value is the predicted signal value memory 31
If it is not stored in the selection circuit 36, the control circuit 35 selects the selection circuit 36.
The prediction circuit 37 controls the prediction circuit 37 based on the previous reproduction signal value 42.
The predicted signal value 43 outputted by is selected.

If an error occurs in the difference information, in the conventional method, the predicted signal value is determined based on all past difference information, so no matter how accurate the subsequent difference information is, the future predicted signal value will be This will be different from the one on the encoding side, and all subsequent reproduced signal values will be incorrect. In the method of the present invention, some predicted signal values are transmitted from the encoding side, so errors in the reproduced signal value do not occur until the next predicted signal value transmitted from the encoding side is used. Unlike the conventional method, all subsequent reproduced signal values will not be incorrect. For example, if one out of 100 predicted signal values generated on the encoding side is transmitted at equal intervals, the error in the reproduced signal value will be 10 consecutive
The number will be less than 0.

〔Effect of the invention〕

As explained above, the present invention transmits some predicted signal values in addition to the difference information from the encoding side to the decoding side, so that even if an error occurs in the difference information on the transmission path, the decoding side reproduces it. This means that errors in the transmitted signal will only occur within a certain period of time, which has the effect of significantly improving transmission quality.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the encoding side of the differential pulse encoding method of the present invention, and FIG. 2 is a block diagram of the decoding side.

Claims (1)

[Claims] (1) In a differential pulse encoding method that encodes a difference value between an input signal value and a predicted signal value, the encoding side has means for transmitting some predicted signal values in addition to the difference information, and the decoding When the coding side adds the difference value to the predicted signal value and reproduces the signal, if the prediction signal value is transmitted from the coding side, the difference value is added to this predicted signal value, and the predicted signal value is A differential pulse encoding method characterized by having means for adding a differential value to a predicted signal value based on past differential values when the signal is not transmitted.