TWI469135B - Adaptive differential pulse code modulation (adpcm) encoding and decoding method - Google Patents

Description

Adaptive differential pulse code modulation coding and decoding method

The invention relates to a method for adapting differential pulse code modulation coding and decoding, in particular to a situation that can avoid the dramatic change of the speech data solved by the decoding end caused by the instability of the wired or wireless channel, so as to enable An innovative designer who effectively reduces the noise and severe distortion of the speech data decoding output, and adjusts the adaptive differential pulse code modulation coding and decoding method in its overall implementation.

According to the wireless voice transmission, the voice data is transmitted on the 2.4GRF with low frequency and low cost, so that the data compression and decompression are required; among them, the commonly used adaptive differential pulse code modulation [Adaptive Differential Pulse Code] Modulation; ADPCM] Encoder (3), please refer to the figure of the existing encoder architecture shown in Figure 5, the encoder (3) mainly includes a subtractor (31), a quantizer (32), and an inverse quantizer ( 33), the predictor (34), the step adaptor (35) and the adder (36), such that the input signal is subtracted from the magnitude of the next data estimated by the predictor (34) by the subtractor (31) After that, an error value is generated. Ideally, the error value should be only a small value, and the error value is further fixed by the quantizer (32) with the step adaptor (35). Under the resolution condition, it is quantized to generate a compressed output code, and the inverse quantizer (33) receives the pressure. After the output code is reduced, the phase synchronization value and the same resolution condition provided by the step adaptor (35) are inversely quantized to send a signal, and then passed through the adder (36) and the predictor (34). The feedback values of the loopback are added to generate the input signal of the predictor (34), and then the steps of subtracting the input signal from the input signal and quantizing and dequantizing the error value are repeated until the end.

However, the above-mentioned adaptive differential pulse code modulation encoder can achieve the expected effect of compressing and encoding the speech data, but it is also found in the overall implementation that the effect of amplitude retention is caused by the decompression of ADPCM. The volume amplitude of the music will remain at the previous amount when decompressed in the next music frame, so it will remain loud for a period of time, or keep a small sound for a period of time, causing the prediction value error when the receiver is decompressed too much. However, problems such as large, small or popping sounds, and murmurs occur when the volume of the sound occurs, so that there is still considerable room for improvement in the overall coding design.

The reason is that the inventor has been rich in design and development and practical production experience of the relevant industry for many years, and has made research and improvement on the existing ADPCM codec method, and provides a method of adaptive differential pulse code modulation and decoding to achieve better and practical. The purpose of value.

The method for adapting differential pulse code modulation coding and decoding according to the present invention is characterized in that the encoder causes the input signal and the magnitude of the next data estimated by the predictor to be subtracted by the subtractor to generate an error value, and the error value is obtained. Then, the quantizer generates an adaptive differential pulse code modulation compression coding signal, which is mainly connected to the quantization step unit by the quantizer, and uses the encoder step index table to perform pre-decompression with the encoder step table. Output the step index value [index] to the delay register of the previous signal step index value, and the delay index of the step index value to the previous signal step index value is delayed by one time for the next adaptation. When the differential pulse code modulation compression coded signal output is output, the previous signal step index value is synchronously output; the above coded output signal is passed through the inverse quantizer of the decoder and a decoder step index table, in the decoder step index table, generating an exponential input multiplexer and performing multiplex processing on the input previous step index value, and providing the step index value in the decoder step table After the vectorizer performs inverse quantization, a signal is sent to the predictor for decompression and decoding operations; thereby, the decoding end caused by the instability of the wired or wireless channel can be avoided, and the decoded speech data is severe The change occurs to effectively reduce the noise and severe distortion of the speech data decoding output.

According to the above concept, the method for adapting the differential pulse code modulation and decoding according to the present invention mainly comprises the following steps: the encoder inputs the original pulse code modulation (PCM) data, and then outputs the adaptive differential pulse of the current audio frame. The code modulation compression code value (ADPCM CODE), the previous index of the previous audio frame and the control information are combined into a single packet, and transmitted to the decoder for decoding by wired or wireless communication. The original pulse code modulation (PCM) data is restored.

According to the above concept, the method for adapting differential pulse code modulation coding according to the present invention comprises the following steps: the encoder compresses and encodes the input data to generate an encoded signal; and the quantization step unit simultaneously receives the differential signal for pre-solution. Compression, the encoder step index table is matched with the encoder step table to generate a one-step index value to the delay register of the previous signal step index value; the delay register delay of the previous signal step index value is one Unit time, output step index value; generate the adaptive differential pulse code modulation compression coded signal, and the encoder of the previous signal step index value, and transmit the signal to the decoder.

According to the above concept, the method for adapting the differential pulse code modulation and decoding according to the present invention is to load the last step index value of the previous frame for decoding when the first audio PCM is restored. The order index value can continue the step index value generated by the last audio PCM of the previous frame, thereby avoiding the decoding end The voice data is changed drastically to maintain the decoded sound quality.

(1)‧‧‧Encoder

(11)‧‧‧Subtractor

(12) ‧ ‧ Quantizer

(13)‧‧‧Quantitative step unit

(131)‧‧‧Encoder step index table

(132)‧‧‧Encoder step scale

(14)‧‧‧Reverse Quantizer

(15)‧‧‧ predictor

(16) ‧‧‧ Delay register of the previous signal step index value

(2)‧‧‧Decoder

(21)‧‧‧Reverse Quantizer

(22)‧‧‧Decoder step index table

(23) ‧‧‧ predictors

(24) ‧‧‧Multiplexer

(25)‧‧‧Decoder step scale

(3)‧‧‧Encoder

(31)‧‧‧Subtractor

(32) ‧ ‧ Quantizer

(33)‧‧‧Reverse Quantizer

(34) ‧‧‧ predictors

(35) ‧ ‧ step adaptor

(36)‧‧‧Adder

First Picture: Overall Architecture Diagram of the Invention

Second picture: encoder architecture diagram of the present invention

Third: Schematic diagram of the encoder transmission signal of the present invention

Fourth figure: decoder architecture diagram of the present invention

Figure 5: Existing encoder architecture diagram

For a more complete and clear disclosure of the technical content, the purpose of the invention and the effect achieved by the present invention, it will be explained in detail below, and please refer to the illustrated figure and figure number:

First, referring to the first figure, the overall architecture diagram of the present invention is shown. The present invention mainly enables the encoder [ADPCM Encoder] (1) to input the adaptive differential pulse code modulation compression to the decoder [ADPCM Decoder] (2). The encoded signal and the previous signal step index value [Previous index]; wherein: the encoder [ADPCM Encoder] (1), please refer to the second diagram of the encoder architecture diagram of the present invention, which is mainly used in one The subtractor (11) is connected to a quantizer (12), and the quantizer (12) is connected with a quantization step unit (13) and an inverse quantizer (Inverse Quantizer) (1). 4), the quantization step unit (13) is connected to an encoder step index table (131) at the back end of the quantizer (12), and the encoder step index table (131) has a back end, respectively The delay register (Previous Index Delay Holder (Z-1)) (16), which is coupled to itself, is connected with an encoder step table (132) and a previous signal step index value (16), the code The step table (132) is respectively connected to the quantizer (12) and the inverse quantizer (14), and the inverse quantizer (14) is connected with a predictor (15), and the predictor (15) is also connected. Connect to the subtractor (11).

In this way, when the present invention is used for compression coding, the pulse code modulation (PCM) input signal and the predictor (15) are estimated, and the size of the next data is estimated to be subtracted by the subtractor (11). An error value is generated, which is quantized by the quantizer (12) to generate an adaptive differential pulse code modulated compression code value (ADPCM CODE), and the encoder step is used in the quantization step unit (13). The index table (131) cooperates with the encoder step table (132) to receive the differential signal for pre-decompression processing, and outputs an index value [Index] from the encoder step index table (131) to the previous signal step. The index value delay register (16) delays one unit time to synchronously output the previous signal step index value and the inverse vector when the next adaptive differential pulse code modulation compression code value [ADPCM CODE] is output. After receiving the adaptive differential pulse code modulated compression coded signal, the chemist (14) will use the encoder step index table (131) to match the encoder step The phase synchronization index value provided in Table (132) is inverse quantized to send a signal to the predictor (15) to generate an input signal of the predictor (15), and then repeating the subtraction from the input signal, Error value quantization and inverse quantization verification steps until the end.

For the signal outputted by the encoder (1), please refer to the third diagram of the encoder transmission signal of the present invention, which can include the adaptive differential pulse code modulation generated by the quantizer (12). Compressing the encoded value, and the delay register (16) of the previous signal step index value, outputting the previous signal step index value and control information; the control information includes the number of each audio frame, the encoder Identification code, packet flow control message and frequency hopping information, etc., and the number of the audio frame can be used by the decoder to count the number of packets lost, so that the number is numbered from 0~n by the encoding end, where n can be a constant More than 100 to 1000 or less; and frequency hopping information, the encoding end can be used to inform the decoder to perform frequency hopping.

In addition, the decoder [ADPCM Decoder] (2), please refer to the fourth diagram of the decoder architecture diagram of the present invention, which is generated by the quantizer (12) of the corresponding encoder (1). The adaptive differential pulse code modulation compression coding signal is respectively provided with an inverse quantizer (Inverse Quantizer) (21) and a decoder step index table (Index table) (22), and the inverse quantizer (21) is connected Predictor (23), the decoder step index table (22) is respectively fed back to itself and connected to the multiplexer [Mux] (24), and a signal preceded by the encoder (1) The step index value of the delay register (16), the output of the previous signal step index value is also connected to the multiplexer (24), and the multiplexer (24) is connected with the decoder step table [Step table] (25), the decoder step table (25) is connected to the inverse quantizer (21).

The adaptive differential pulse code modulated compression encoded signal and the previous signal step index value transmitted by the encoder (1) to the decoder (2), the adaptive differential pulse code modulated compression encoded signal, respectively inputting the decoding The inverse quantizer (21) of the (2) and the decoder step index table (22), in the decoder step index table (22), generate the exponential input multiplexer (24) and the input previous signal step After the multiplex processing of the order index value, when the first audio PCM is restored, the last step index value of the previous audio frame must be loaded for decoding, so that the index value can continue the last audio of the previous audio frame. The step index value generated by the PCM is inversely quantized by the inverse quantizer (21) by providing the step index value by the decoder step table (25), and a signal is sent to the predictor (23) for decompression and Decoding action to maintain the volume is large or small, without the phenomenon of sudden increase and small, thus avoiding the situation that the speech data solved by the decoding end is drastically changed, so as to effectively reduce the noise and severeness of the speech data decoding output. Distortion to maintain the quality of the decoded sound stable.

As described above, the composition and method implementation of the structure of the present invention show that, in comparison with the prior art, the present invention mainly avoids the speech that is decoded by the decoding end due to the instability of the wired or wireless channel. The data has been changed drastically, so as to effectively reduce the noise and severe distortion of the speech data decoding output, which can be multi-party Used in wireless music speakers, digital classroom wireless computer control devices, digital home wireless computer control devices and PUB DJ and remote control.

However, the above-described embodiments or drawings are not intended to limit the structure or the use of the product of the present invention, and any suitable changes or modifications may be made without departing from the scope of the invention.

In summary, the embodiment of the present invention can achieve the expected use efficiency, and the specific method disclosed therein has not been seen in the same product, nor has it been disclosed before the application, and has completely complied with the provisions of the patent law. And the request, the application for the invention of a patent in accordance with the law, please forgive the review, and grant the patent, it is really sensible.

(1)‧‧‧Encoder

(2)‧‧‧Decoder

Claims (8)

A method for adapting differential pulse code modulation and decoding, mainly by encoding the original pulse code modulation (PCM) data after encoding, and outputting the adaptive differential pulse code modulation compression coding value of the current audio frame (ADPCM) CODE), the previous index of the previous audio frame and the control information are combined into a single packet and transmitted to the decoder for decoding to restore the original pulse code modulation (PCM) data. The method for adapting differential pulse code modulation coding and decoding according to claim 1, wherein the encoder and the decoder are transmitted by wired communication. The method for adapting differential pulse code modulation coding and decoding according to claim 1, wherein the encoder and the decoder are transmitted by wireless communication. The method for adapting differential pulse code modulation coding and decoding according to claim 1, wherein the coding method comprises the following steps: the encoder compresses and encodes the input audio pulse code modulation (PCM) data to generate an adaptation. Differential differential pulse code modulated compression coded value (ADPCM CODE); the quantization step unit simultaneously receives the differential signal for pre-decompression to encode The step index table cooperates with the encoder step table to generate a one-step index value to the delay register of the previous signal step index value; the delay register of the previous signal step index value is delayed by one unit time, and the output step The index value is generated by generating the current adaptive differential pulse code modulation compression code value and the encoder transmission signal of the previous signal step index value to the decoder. The method for adapting differential pulse code modulation coding and decoding according to claim 1, wherein the decoder performs audio data on the input adaptive differential pulse code modulated compression code and the step index value during decoding. Restore. The method for adapting differential pulse code modulation coding and decoding according to claim 1, wherein the control information includes a number of each audio frame, an encoder identification code, a packet flow control message, and a frequency hopping information. The method for adapting differential pulse code modulation coding and decoding according to item 6 of the patent application scope, wherein the number of the audio frame is used by the decoder to count the number of packets lost. The method for adapting differential pulse code modulation coding and decoding according to item 6 of the patent application scope, wherein the frequency hopping information is used by the coding end to inform the decoding end to perform frequency hopping.