KR20040050810A - Method for restoring packet loss by using additional speech data and transmitter and receiver using the method - Google Patents

Abstract

PURPOSE: A method for recovering packet loss by using additional voice data is provided to use voice packet data as additional information for an FEC(Forward Error Correction), and to use additional voice packet data for an error correction in case a packet loss occurs, then to use an error hiding method in case the packet loss consecutively occurs, thereby improving deteriorated sound quality caused by the packet loss. CONSTITUTION: A voice signal is inputted in 1 frame unit(S61). A system decides whether voice signals of all frames are inputted(S62). If not, the system starts coding(S63). The coding process is divided into a low bit rate coding(S64) and a high bit rate coding(S66). The low bit rate coding is performed at 6.6kbit/s bit rate, and a voice signal coded at low bit rate is delayed for 1 frame(S65). A voice signal coded at high bit rate and the voice signal coded at the low bit rate are synthesized into one packet(S67). If the voice signals of all the frames are inputted, the system transmits the packets generated before this time(S68).

Description

Packet Loss Recovery Using Additional Voice Data and Transceiver Using Them {Method for RESTORING PACKET LOSS BY USING ADDITIONAL SPEECH DATA AND TRANSMITTER AND RECEIVER USING THE METHOD}

The present invention relates to a packet loss recovery method using additional voice data and a transceiver using the same. More particularly, the low-rate voice packet data is used as additional voice data for packet loss frequently occurring in packet communication such as an Internet environment. And a transceiver using the method.

Due to the high propagation delay and packet loss, which is characteristic of packet-switched networks, the Internet is very difficult to implement real-time voice communication. In order to enable real-time multimedia communication services in the Internet in such an environment, voice coding technology, compression technology, packet recovery, and error concealment technology are required.

In Fig. 1, a technique for recovering voice packet loss is classified, and this classification is performed according to a sender based recovery method. As shown in FIG. 1, a packet loss recovery technique is largely divided into an active retransmission method and a passive channel coding method. In the active retransmission method, when a packet loss is confirmed at the receiving side, the transmitting side retransmits a portion corresponding to the lost packet. The passive channel coding is further divided into channel coding techniques including traditional forward error correction (FEC) and interleaving based methods. The additional information in the forward error correction is typically a media-independent form based on an exclusive OR (XOR) operation, and media-specific using the characteristics of the transmitting medium. It is divided into forms again.

Among the above methods, the active retransmission method has high data reliability but needs feedback from the receiving side to the transmitting side, which causes delay due to feedback, which is suitable for implementing a real-time multimedia service using this method. Not. In the case of the above-described passive channel coding method, it is possible to reduce packet loss by distributing the loss in the interpolation method, but the delay increases like the active retransmission method. Recovering lost packets using forward error correction by medium independent side information can process and recover lost packets regardless of the contents of the packet, and there is a small amount of computation required to induce error correction packets. There is also an additional delay.

Therefore, the lost packet recovery method for implementing the real-time voice communication service does not require feedback information, and it is required to perform a forward error correction function using an existing speech encoding algorithm. In this case, the structure of the overall voice communication system is simplified, and since the forward error correction code based on the used voice packet data is composed of important feature data of voice, it is possible to maintain good sound quality during voice synthesis.

The broadband adaptive multirate voice encoding algorithm is a multirate voice encoding algorithm that operates at nine rates, and is designed to transmit high quality sound quality, and can be used in a mobile communication system or an Internet telephone. In the present invention, when there is no packet loss, it operates like an existing encoder without additional information. When packet loss occurs, the voice packet data of the low rate speech encoder is forwarded by using the multi-rate feature of the broadband adaptive multi-rate encoder. It is used as an error correction code and additionally transmitted to existing voice data, and the receiving end is characterized in that the forward error correction code is used to recover lost packets. In this case, since the bandwidth is increased, it is preferable to select 6.60 Kbit / s or 8.85 Kbit / s, which is a low data rate, for the additional information of the encoder.

On the other hand, as a prior patent on the recovery of packet loss, Korean Patent Registration No. 341401 (Registration Date: June 7, 2002) discloses "Adaptive additional transmission method and packet loss recovery method for interactive audio service and multimedia for the same. Audio input / output control device of computer "is known. The prior patent has an additional transmission processor composed of a transmitting side and a receiving side, and the transmitting side dynamically sets the additional transmission scheme using the loss information fed back from the receiving side, and the receiving side recovers the lost packet using the additional data. do. In the preceding patent, since the loss information is required to be fed back from the receiver to configure additional information, there is a problem in that a delay caused by the feedback is inevitably caused, and the loss rate is increased by changing the delay degree of the additional information according to the loss rate. As a result, the decoding delay increases.

In addition, as a prior patent relating to the recovery of packet loss, Korean Patent Registration No. 238324 (Registration Date: October 13, 1999) discloses a method and apparatus for error concealment of an audio signal. The prior patent consists of a triangular converter, a selector, and an error concealment data generator, and is characterized by reconstructing the audio data by finding a conversion coefficient that best characterizes the audio signal component of the frame before and after the lost frame using the concentration. However, the above-mentioned patent has a problem in that the accuracy of recovery is inferior because the error is concealed by interpolating the frame data before and after the lost frame when data is lost, and additional information is not used.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described technical problems. The present invention encodes an input speech into high bit rate information and a low bit rate information, and configures and transmits the low bit rate information as additional speech data, and packet loss at the receiving end. In this case, the lost packet is recovered by using the additional voice data, while when two or more consecutive packets are lost, one packet is recovered by using the additional voice data and the remaining packets are future and past packets. An object of the present invention is to provide a packet loss recovery method for recovering by applying interpolation between parameters, and a transceiver to which the recovery method is applied.

1 is a diagram of a technique for recovering voice packet loss.

2 is a diagram illustrating a structure of a transceiver to which a packet loss recovery method according to the present invention is applied.

3 is a diagram illustrating a principle of recovering packet loss using additional voice data.

4A and 4B are diagrams illustrating bit allocation per mode in a wideband adaptive multiple bit rate (AMR-WB) encoder.

5 is a diagram illustrating a lost packet reproduction process using interpolation.

6 is a diagram illustrating a procedure at a transmitting end of a lost packet recovery method according to the present invention.

7 is a diagram illustrating a procedure at a receiving end of a lost packet recovery method according to the present invention.

(Explanation of symbols for the main parts of the drawing)

11: voice input unit 12: encoder

121: low bit rate encoder 122: high bit rate encoder

13 delay unit 14 packet configuration unit

15: packet transmission unit 20: the Internet

31: packet receiving unit 32: frame deletion / hiding unit

33: buffer 34: encoder

341: speech encoder 35: speech output unit

A packet loss recovery method at a transmitting end using additional voice data according to an aspect of the present invention for achieving the above object,

Receiving a voice signal in units of frames;

A second step of encoding the voice signal input in the first step at a low bit rate and a high bit rate, respectively;

A third step of delaying one frame of the speech signal encoded at a low bit rate in the second step;

A fourth step of composing a packet by synthesizing a signal encoded at a low bit rate with one frame delay obtained in the third step and a signal encoded at a high bit rate obtained in the second step; And

And a fifth step of transmitting the packet obtained in the fourth step when the processes of the first to fourth steps are completed for all voice data.

In addition, the packet loss recovery method at the receiving end using the additional voice data according to another aspect of the present invention,

A first step of receiving packet data and receiving the data in the form of a bit stream;

A second step of checking whether a loss occurs in a current frame of the packet data input in the first step;

A third step of copying correctly received frame data immediately before the loss occurs to the lost frame data if the number of accumulated lost frames is greater than or equal to a predetermined value when the loss occurs in the second step;

A fourth step of checking the accumulated number of lost frames if no loss occurs in the second step;

When the number of lost frames accumulated in the fourth step is "0", the data encoded at the high bit rate among the received packet data is replaced with the correctly received data immediately before the packet loss occurs, and the data encoded at the high bit rate A fifth step of decoding the original speech signal;

When the number of lost frames accumulated in the fourth step is "1", the additional loss data encoded at a low bit rate added to a packet received next to the lost frame is decoded to recover the lost frame. A sixth step of decoding an encoded data at a high bit rate to generate an original speech signal; And

When the number of lost frames accumulated in the fourth step is "2", the interpolation method is applied to the additional voice data for the first lost frame and the correctly received data immediately before the loss. Data is recovered, and the remaining frame includes the seventh step of recovering by applying the procedure of the sixth step, and generating the original speech signal by decoding the data encoded at the high bit rate.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2 shows a structure of a transceiver to which a packet loss recovery method according to the present invention is applied.

The transmitter shown in FIG. 2 is a transmitter comprising a voice input unit 11, an encoder 12 according to the G.722.2 standard, a delay unit 13, a packet constructing unit 14, and a packet transmitting unit, and a packet receiving unit 31. ), A frame deleting / hiding unit 32, a buffer unit 33, a decoder 34 and a voice output unit 35 according to the G.722.2 standard. 20) are connected to each other. The encoder 12 includes a low bit rate encoder 121 and a high bit rate encoder 122, and the decoder 34 includes a voice decoder 341. The encoder 12 is an encoder according to the G.722.2 standard and is a wideband adaptive multiple bit rate (AMR-WB) encoder.

When the operation is started, the voice signal input from the voice input unit 11 of the transmitter is input to the encoder 12, and the voice signal is transmitted to the low bit rate encoder 121 and the high bit rate encoder 122 of the encoder 12. Are respectively encoded. The speech signal output from the low bit rate encoder 121 is delayed for one frame in the delay unit 13 for use as additional information of the next packet. The outputs of the delayer 13 and the high bit rate encoder 122 are input to the packet constructing unit 14, and the packet constructing unit 14 is a high bit rate of the current frame output from the high bit rate encoder 122. The encoded speech signal and the speech signal encoded at the low bit rate of the previous frame output from the delay unit 13 are combined into one packet. The speech signal encoded at the low bit rate is additional speech data used to recover a packet lost at the receiving end. The packet data obtained by the packet configuring unit 14 is transmitted to the packet communication network such as the Internet 20 via the packet transmitting unit 15.

Next, the operation in the receiver will be described. The packet receiving unit 31 of the receiver receives packet data from a packet communication network such as the Internet 20, and the received packet data is input to the frame erasing / hiding unit 32 and the buffer 33, respectively. The frame deletion / hiding unit 32 checks whether the packet data of the input frame has been normally received. If a loss occurs in one packet, the frame deletion / hiding unit 32 recovers the lost packet by using additional voice data, and consecutive two or more packets. When a loss occurs in the first packet, the lost first packet recovers the lost packet using additional voice data, and the remaining packets are recovered by applying interpolation to the previously correctly received packet and the recovered first packet. Next, the voice packet data output from the frame deletion / hiding unit 32 is decoded by the voice decoder 341 in the decoder 34, and the decoding is performed according to the G.722.2 standard. The signal decoded by the decoder 34 is input to the voice output unit 35, and the signal decoded by the voice output unit 35 is converted into a voice signal and output.

3 illustrates a principle of recovering packet loss using additional voice data. As shown in FIG. 3, each packet is composed of encoded speech packet data of a current frame and additional speech data of a previous frame. That is, the n th packet is composed of n th voice packet data and additional voice data of the (n-1) th packet. Each packet of Fig. 3 is composed of two blocks, the first block representing additional speech data of the previous frame and the second block representing encoded speech packet data of the current frame.

As described above, if a packet received at the receiver is lost, and a loss occurs in a single packet, for example, the (n + 1) th packet, it is added to the (n + 2) th packet immediately after the packet. The (n + 1) th packet is recovered by the additional audio data. In case of loss of two consecutive packets, for example, the nth packet and the (n + 1) th packet, the receiver adds the (n + 1) th packet included in the (n + 2) th packet first. Recover the (n + 1) th packet with voice data and interpolate the last packet correctly received before packet loss occurred, i.e., the (n-1) th packet and the recovered (n + 1) th packet Apply n to reproduce the n th packet. Meanwhile, when a loss occurs in two or more packets, for example, when a loss occurs in the (n-1) th packet of FIG. 3, a method for copying a previously received packet, that is, the (n-2) th packet, is copied. This can be used. In this case, in order to reduce propagation of an error due to asynchronous state, which is a characteristic of an encoder based on Code Excited Linear Prediction (CELP), the degradation of sound quality can be reduced by initializing the past excitation signal.

The transceiver according to the present invention utilizes a multi-rate, which is a feature of a wideband adaptive multi-rate coder. In other words, instead of using two different voice coders, a voice signal is encoded into high bit rate information and low bit rate information using a wideband adaptive multiple bit rate (AMR-WB) coder having several modes. That is, the transmitter generates and transmits one packet by mixing an original speech signal composed of high bit rate information and additional speech data composed of low bit rate information for one frame. The receiver recovers the original speech signal by decoding using the high bit rate information, and when a packet loss occurs, forward error correction (FEC) is performed using additional speech data added to the next packet and transmitted. Recover lost packets by doing 4A and 4B are diagrams showing bit allocation for each mode in a wideband adaptive multiple bit rate (AMR-WB) encoder. In the transceiver according to the present invention, when the loss does not occur, the mode of the wideband adaptive multi-rate encoder is used in the same manner, and when the packet loss occurs and the additional information is inserted, the transmission rate of the additional information is selected and used. . For example, in the transceiver according to the present invention, 6.6 kbps mode may be used as additional information and a total of 16 transmission modes may be configured as follows. In order to indicate 16 modes, a total of 4 bits are allocated to use the nine modes of the broadband adaptive multi-rate encoder from '0000' (0) to '1000' (8), and from '1001' (9) to ' The seven modes up to 1111 '(15) are modes in which additional information is included in the wideband adaptive multiple bit rate encoder (modes 1 to 8).

In the present invention, when two or more consecutive packets are lost, the first packet is recovered using additional speech data using the feature of the speech encoder, and the remaining packets are reproduced using interpolation to recover the lost packets. A concealment technique is used.

Currently, the best packet loss recovery method among speech encoders is to use a lost packet by attenuating the parameters of previously received packets. This method is used in recent speech coder standards because it has the advantage of not requiring additional delay. However, since the parameter values are uniformly attenuated, sound distortion occurs seriously when loss occurs in consecutive packets.

The lost packet recovery method according to the present invention is a method using additional voice data, and can be effectively applied when loss occurs in a non-recoverable continuous packet as shown in FIG. 5. That is, as shown in Fig. 5, when a loss occurs in two consecutive packets, one packet can be recovered as additional voice data, but the other packet cannot be recovered by the existing technology. In the present invention, as shown in Fig. 5, the packet recovered by the additional voice data is used as future packet information, and the packet correctly received immediately before the non-recoverable packet is used as the past packet. By applying interpolation between the parameters of, it is possible to generate a packet that can replace the non-recoverable packet. Here, linear interpolation is applied to the adaptive codebook gain, and median smoothing is applied to the adaptive codebook index. However, since it is difficult to apply interpolation for LSF and fixed codebook gain, the parameters of future packets are copied as is. If a loss occurs in three or more consecutive packets, the last correctly received packet before the loss occurs is copied and used as a replacement packet of the first lost packet.

In general, an encoder based on Code Excited Linear Prediction (CELP) generates a model parameter based on the current state of speech. To transmit and decode these parameters, it creates its own state to play the voice. In a general operation, since the change in the state updated while the encoder is updated is the same as the change in the state updated while the decoder is decoding, the decoder can reproduce the original sound from the transmitted parameter. However, if packet loss occurs, the state synchronization between the encoder and the decoder no longer matches, and the decoder cannot smoothly update the state. Therefore, even if the packet loss no longer occurs, the sound quality inevitably occurs because the decoder reproduces even the packet which is not correctly received using the state that is not correctly updated. As a result, an incorrect prediction state causes a deterioration in sound quality when decoding a signal after a lost frame, so that the speech signal should not be recovered without correctly updating the prediction state. In the lost packet recovery method according to the present invention, a state error can be reduced by using a method of initializing a past excitation signal to "0" for a frame that the decoder cannot recover.

Next, the procedure at the transmitting and receiving end of the lost packet recovery method according to the present invention will be described with reference to FIGS. 6 and 7. 6 shows a procedure at a transmitting end of a lost packet recovery method according to the present invention, and FIG. 7 shows a procedure at a receiving end of a lost packet recovery method according to the present invention.

First, the procedure at the transmitting end of the lost packet recovery method according to the present invention will be described with reference to FIG.

When the operation starts, the voice signal is input in units of one frame (S61). Next, it is determined whether the audio signals of all the frames are input (S62). If the voice signals of all the frames are input in the step S62, the already generated packet is transmitted (S68). If the voice signals of all the frames are not input in step S62, encoding, that is, encoding is started (S63). In encoding, low bit rate encoding (S64) and high bit rate encoding (S66) are performed in parallel. The low bit rate encoding is performed at a bit rate of 6.6 kbit / s, and an operation S65 is delayed for one frame for the speech signal encoded at the low bit rate. The speech signal on which the high bit rate encoding is performed and the speech signal on which the low bit rate encoding is delayed by one frame are synthesized into one packet (S67). At this time, one packet includes a speech signal encoded at the high bit rate of the current frame and additional speech data encoded at the low bit rate of the previous frame. When all the frames of the voice signal are input in step S62, the packets generated during the transmission are transmitted (S68).

Next, the procedure at the receiving end of the lost packet recovery method according to the present invention will be described with reference to FIG.

When the operation of the receiver is started, packet data is received from a packet communication network such as the Internet (S71). Next, the received packet data is input in the form of a bit stream (S72). Next, it is determined whether all received data is input (S73), and when all data is input, the packet recovery procedure ends. If all data is not input in step S73, the bad frame indicator variable (BFI: Bad Frame Indicator) is inspected to determine whether a loss occurs in the packet (S74). The bad frame indication variable has a state value of "0" or "1", and if the loss occurs in the frame, the state value is "0", and if there is no loss in the frame, the state value is "1". ". Next, it is determined whether the bad frame indication variable BFI is "0", that is, whether there is a loss in the frame (S75).

When there is a loss in the current frame in step S75, that is, when the bad frame indication variable BFI is "0", it is determined whether the bad frame count variable (BFC: Bad Frame Count) is "3" (S76). ). If the bad frame count variable is not "3" in step S76, the bad frame count variable is counted up one step (S79), and the process jumps to the step S72. On the other hand, if the bad frame count variable BFC is "3" in step S76, R-data (the concept defined in the embodiment of the present invention) is the last frame data correctly received before packet loss occurs. Is decoded by copying the frame into the lost frame (S77), and the bad frame count variable BFC is down counted (S78).

If there is no loss in the current frame in step S75, that is, if the bad frame indication variable BFI is "1", it is determined whether the bad frame count variable BFC is one of "0, 1, 2". (S80, S81). When the bad frame count variable BFC is "0" in the step S80, there is no loss in the current frame as well as no accumulated loss frames. In this case, encoding is performed at a high bit rate among the received packet data. The received data H-data is replaced with the last frame data R-data correctly received before packet loss occurs (S83). Next, the original audio signal is generated by decoding the H-data encoded at the high bit rate (S85), after initializing the bad frame count variable (BFC) to "0" (S72). Jump to

If it is determined in step S81 that the bad frame count variable BFC is "1", this case is a case where a loss occurs in one input frame. Accordingly, the frame in which the loss occurs is recovered by decoding the additional speech data encoded at a low bit rate of 6.60 kbit / s added to the packet after one frame (S84). Next, the normal encoding process is performed by passing the above-described step (S85).

If it is determined in step S81 that the bad frame count variable BFC is "2", this case is a case where a loss occurs in two consecutive input frames. In this case, interpolation is performed on the additional speech data for the first lost frame encoded at a low bit rate of 6.6 kbit / s and the correctly received data (R-data) immediately before the loss. Data is recovered (S82). Next, after the loss of the remaining frames are recovered through the step S84, the normal decoding process is performed by the step S85. By doing so, appropriate recovery can be made according to the number of lost frames, and normal decoding can be performed.

While the invention has been described above based on the preferred embodiments thereof, these embodiments are intended to illustrate rather than limit the invention. It will be apparent to those skilled in the art that various changes, modifications, or adjustments to the above embodiments can be made without departing from the spirit of the invention. Therefore, the protection scope of the present invention will be limited only by the appended claims, and should be construed as including all such changes, modifications or adjustments.

As described above, the present invention recovers one packet loss by using a forward error correction (FEC) method by using the multiple bit rate of the broadband adaptive multiple bit rate (AMR-WB) encoder, and correctly receives the unrecoverable packet in the past. Packet loss is recovered by concealing errors by applying interpolation to packets recovered from one packet and additional information. In addition, in order to reduce asynchronous state error caused by packet loss, the decoder can initialize the past excitation signal to '0' to reduce the state error. Since this method does not require feedback information and does not cause delay, it can be used as a transmission loss compensation method of a broadband adaptive multi-rate encoder for voice communication between the IMT-2000 system and the Internet.

Claims (12)

Receiving a voice signal in units of frames; A second step of encoding the voice signal input in the first step at a low bit rate and a high bit rate, respectively; A third step of delaying one frame of the speech signal encoded at a low bit rate in the second step; A fourth step of composing a packet by synthesizing a signal encoded at a low bit rate with one frame delay obtained in the third step and a signal encoded at a high bit rate obtained in the second step; And And a fifth step of transmitting the packet obtained in the fourth step when the processes of the first to fourth steps are completed for all voice data. Packet loss recovery method in a transmitting end using additional voice data. The method of claim 1, The low bit rate encoding in the second step is characterized by consisting of a bit rate of 6.6 kbit / s Packet loss recovery method in a transmitting end using additional voice data. The method of claim 1, The speech signal encoded at the low bit rate in the third step is used as additional speech data, and is delayed by one frame and combined and transmitted in a packet together with the speech signal of the next frame. Packet loss recovery method in a transmitting end using additional voice data. The method of claim 1, The low bit rate and high bit rate encoding is performed in a wideband adaptive multi-rate encoder. Packet loss recovery method in a transmitting end using additional voice data. A first step of receiving packet data and receiving the data in the form of a bit stream; A second step of checking whether a loss occurs in a current frame of the packet data input in the first step; A third step of copying correctly received frame data immediately before the loss occurs to the lost frame data if the number of accumulated lost frames is greater than or equal to a predetermined value when the loss occurs in the second step; A fourth step of checking the accumulated number of lost frames if no loss occurs in the second step; When the number of lost frames accumulated in the fourth step is "0", the data encoded at the high bit rate among the received packet data is replaced with the correctly received data immediately before the packet loss occurs, and the data encoded at the high bit rate A fifth step of decoding the original speech signal; When the number of lost frames accumulated in the fourth step is "1", the additional loss data encoded at a low bit rate added to a packet received next to the lost frame is decoded to recover the lost frame. A sixth step of decoding an encoded data at a high bit rate to generate an original speech signal; And When the number of lost frames accumulated in the fourth step is "2", the interpolation method is applied to the additional voice data for the first lost frame and the correctly received data immediately before the loss. A seventh step of recovering data, restoring the remaining frames by applying the sixth step, and decoding the encoded data at the high bit rate to generate an original speech signal. Packet loss recovery method at the receiving end using additional voice data. The method of claim 5, In the first step, the packet data is received from a packet communication network such as the Internet. Packet loss recovery method at the receiving end using additional voice data. The method of claim 5, In the fifth to seventh step, the past excitation signal is initialized to reduce the state error after the recovery of the lost frame. Packet loss recovery method at the receiving end using additional voice data. The method of claim 5, In the third step, the predetermined value is a case where the number of cumulative lost frames is three. Packet loss recovery method at the receiving end using additional voice data. A voice input unit receiving a voice signal in units of frames; An encoder for encoding a speech signal input through the speech input unit according to the standard of G.722.2, wherein the encoding is performed in parallel at a low bit rate and a high bit rate; A delayer for delaying a low-rate encoded speech signal for one frame by the encoder; A packet constructing unit configured to combine a speech signal encoded at a low bit rate by the encoder as additional speech data and to combine the speech signal encoded at a high bit rate of a next frame to form one packet; And And a packet transmitter for transmitting the packet data obtained by the packet configuring unit to a packet communication network. Transmitter with packet loss recovery method. The method of claim 9, The low bit rate encoding performed by the encoder is characterized in that the bit rate of 6.6 kbit / s Transmitter with packet loss recovery method. A packet receiver for receiving packet data from a packet communication network; A buffer unit for temporarily storing the packet data received by the packet receiving unit; A frame deletion / hiding unit which checks packet data stored in the buffer unit on a frame-by-frame basis to determine whether to lose or not, and restores the lost frame by using additional voice data and interpolation according to the number of lost frames; A decoder which decodes the voice data of each frame according to the standard of G.722.2; And A voice output unit configured to output a voice signal decoded by the decoder; Receiver with packet loss recovery method. The method of claim 11, The frame erasing / hiding unit recovers a lost frame by using the additional voice data when a loss occurs in one frame, and when the loss occurs in two or more frames, the first frame that is lost uses the additional voice data. And recovers lost frames, and recovers the remaining lost frames by applying interpolation to the previously correctly received frames and the recovered frames. Receiver with packet loss recovery method.