JP2004222150A - IP telephone terminal and data conversion method thereof - Google Patents

Abstract

In an IP telephone terminal, a packet reception interval is calculated, and a D / A conversion unit is driven at a conversion speed in accordance with the reception interval to prevent sound interruption.
In an IP telephone terminal, when a packet arrives, a packet receiving unit transmits a packet arrival trigger to an arrival interval calculating unit. The arrival interval calculation unit 700 reads the current time from the clock 600 at the time of receiving the trigger. Then, the arrival interval calculation unit 700 calculates the arrival interval of the packets by obtaining the difference between the time recorded when the previous packet was received and the current time. A ratio R between the calculated arrival interval and the arrival interval specified in the protocol is obtained, and a frequency F ′ obtained by delaying the reproduction frequency F specified in the protocol by R is obtained. The arrival interval calculation section 700 sets this F 'as the conversion speed in the D / A conversion section 400.
[Selection diagram] Fig. 1

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an IP telephone terminal using the Internet and a data conversion method thereof.
[0002]
[Prior art]
In an IP telephone, a jitter absorption buffer is generally used in order to prevent variations in packet arrival intervals (jitter) from affecting voice reproduction.
The audio is converted into a digital code string by an A / D conversion unit, subjected to a process such as compression in an encoding unit, divided into packets, and transmitted from a packet transmission unit. Generally, packet division is performed at substantially equal intervals, and packets are transmitted at substantially equal intervals.
Queues and the like of the transmitted packets occur randomly in the network, and when arriving at the receiving side, the arrival intervals vary.
The packet received by the packet receiving unit is decoded by the decoding unit, and is converted into voice by the D / A conversion unit. In the case of an IP telephone without a jitter buffer, if the arrival interval between packets is long, a sound break occurs during the interval.
[0003]
When a jitter buffer is provided, audio data is temporarily stored in the jitter absorption buffer without immediately starting data conversion after receiving and decoding the first packet. Thus, when packets arrive at the same interval as the transmission interval, the amount stored in the buffer is constant, and when packets arrive at an interval shorter than the transmission interval, the amount stored in the buffer is temporary. To increase. When packets arrive at intervals longer than the transmission interval, the amount stored in the buffer decreases, but no sound interruption occurs as long as there is data in the buffer.
[0004]
In some conventional techniques, the amount of accumulation in the jitter buffer is dynamically changed. In addition, a method of synchronizing the encoding clock and the reproduction clock using the time information and the number of received bytes included in the encoded stream, and a method using the time information included in the MPEG-2 stream for synchronization (for example, There is a technique that synchronizes both transmission and reception clocks by exchanging clock synchronization packets (see, for example, Patent Document 2).
[0005]
[Patent Document 1]
JP-A-8-307863 [Patent Document 2]
JP-A-6-30043
[Problems to be solved by the invention]
The above-described method of preventing sound interruption using the jitter buffer presupposes that the average arrival interval and the reproduction speed are almost equal, although the arrival intervals of the received packets vary temporarily.
If the packet arrival interval continues to be longer than the length of the audio data included in the packet content in a long cycle, the amount of data stored by the jitter absorbing buffer gradually decreases, and the buffer storage amount becomes zero. The sound break occurs at the point when it becomes.
The above phenomenon is caused, for example, when the communication band becomes smaller than the band required for voice transmission for a long period of time, or the processing load on the transmission side increases, and the time required for encoding the audio data represented by the code Occurs when the length becomes longer than the length. The above phenomenon also occurs when the operating frequency of the transmitting side is lower than that of the receiving side due to an error.
Also, when the operating frequency of the transmitting side is faster than that of the receiving side due to an error, the data amount of the jitter absorption buffer gradually increases, and if the area prepared in the buffer is used up, data overflow due to buffer overflow occurs. As a result, sound skips occur.
In addition, RTP (Real Time Transport Protocol) (IETF RFC1889) provides a method for exchanging information for detecting a difference in operating frequency between a transmitting side and a transmitting side using RTCP (Real Time Transport Control Protocol). Is used to calculate the difference between the operating frequency on the transmitting side and the operating frequency on the receiving side, and also requires an operation to remove the effect of fluctuations in transmission delay, so that the processing is complicated.
The present invention has been made in view of such a point, and an IP telephone which calculates a packet reception interval and drives a D / A conversion unit at a conversion speed in accordance with the reception interval to thereby prevent sound cutoff. It aims to provide terminals.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, the IP telephone terminal according to claim 1, reads out the voice packet data from a jitter absorption buffer for temporarily storing received voice packet data, and converts the voice packet data into an analog voice signal. In the IP telephone terminal provided with the unit, the reception time of the voice packet data is obtained, and the reception interval of the voice packet data is calculated from the reception time of the latest voice packet data and the reception time of the previous voice packet data. Calculating means for calculating the conversion rate of the D / A conversion section based on the reception interval, and controlling the D / A conversion section to convert the latest voice packet data at the conversion rate. And characterized in that:
[0008]
According to a second aspect of the present invention, there is provided a data conversion method for an IP telephone terminal comprising a D / A converter for reading out the voice packet data from a jitter absorption buffer for temporarily storing received voice packet data and converting the voice packet data into an analog voice signal. A first step of obtaining a reception time of the audio packet data, calculating a reception interval of the audio packet data from a reception time of the latest audio packet data and a reception time of the previous audio packet data; Calculating a conversion speed of the D / A conversion unit based on the reception interval, and controlling the D / A conversion unit to convert the latest voice packet data at the conversion speed. It is characterized by the following.
[0009]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a schematic configuration of a receiving block 10 of an IP telephone 1 according to an embodiment of the present invention, and FIG. FIG. 1 is a block diagram illustrating a schematic configuration of an IP telephone 1 that uses 729 as an encoding method.
In FIG. 1, a receiving block 10 reproduces a received audio signal packet and reproduces it as a sound wave. The receiving block 10 includes a packet receiving unit 100, a decoding unit 200, a jitter absorption buffer 300, a D / A conversion unit 400, and a speaker 500. The packet received by the packet receiving unit is decoded by the decoding unit 200, and the data is temporarily stored in the jitter absorption buffer 300, then converted into audio by the D / A conversion unit 400, and output from the speaker 500.
Reference numeral 600 denotes a clock for measuring a packet reception time, and 700 denotes an arrival interval calculation unit (calculation means, control means) for measuring a packet reception interval.
When the packet arrives, the packet receiving unit 100 transmits a packet reception trigger to the arrival interval calculating unit 700 and passes data to the decoding unit 200.
[0010]
FIG. 3 shows the configuration of the arrival interval calculation unit 700. The clock reading unit 710 of the arrival interval calculation unit 700 reads the current time from the clock 600 by a trigger that has received a packet. Then, the difference between the previous packet reception time previously recorded in the previous time storage unit 720 (storage means) and the current time is obtained, and the arrival interval between the previously arrived packet and the currently arrived packet is calculated.
Ratio of calculated arrival interval to arrival interval specified in protocol R = (measured arrival interval) / (predetermined arrival interval) (1)
And a frequency F ′ obtained by delaying the reproduction frequency F specified in the protocol by R.
F '= F / R (2)
Ask for.
The arrival interval calculation section 700 sets this F 'in the D / A conversion section 400, and the D / A conversion section 400 reads out the audio data from the jitter absorption buffer 300 at the designated frequency F' and reproduces it as audio.
In the configuration described in the above-mentioned prior art, the D / A converter only operates at a constant frequency specified in the protocol, but the D / A converter of the present invention operates at the reproduction frequency set by the arrival interval calculator. Operate.
[0011]
In FIG. 2, a control block 11 controls the entire operation of the telephone, and furthermore, controls a call control protocol ITU-TH. A message conforming to the H.323 recommendation is exchanged with a telephone or a gatekeeper of a called party to realize connection / disconnection of a call.
The transmission block 12 includes a transmission device 121, an A / D conversion unit 122, a buffer 123, an encoding unit 124, and a packet transmission unit 125, and packetizes the voice input by the transmission device 121 and transmits the packet.
The physical layer transmission / reception unit 13 is used in common by all functions of the reception block 10, the control block 11, and the transmission block 12, and performs connection to the Internet and transmission / reception of packets.
[0012]
Next, the operation of the present invention will be described.
The sound wave converted to an analog electric signal by the transmitter 121 is converted to a digital signal by the A / D converter 122 at a sampling frequency of 8 kHz. The converted digital signal is stored in the temporary buffer 123.
Encoding section 124 converts the audio signal input from buffer 123 into ITU-T Recommendation G. 729. ITU-T Recommendation G. In 729, encoding is performed in units of one frame with 80 frames of audio for 10 milliseconds as one frame. The encoding unit 124 outputs the code string 10 octets for one frame from the buffer 123 to the packet transmitting unit 125. The packet transmission unit 125 adds an IP header, a UDP header, and an RTP header in order to transmit the encoded code sequence in the IP packet, and transmits the IP packet to the physical layer transmission / reception unit 13. Since the packet is processed and output in units of 10 milliseconds, the processing time of the encoding process is substantially constant, and the packet is transmitted at intervals of 10 milliseconds.
[0013]
The voice packet received by the physical layer transmission / reception unit 13 is received by the packet reception unit 100 of the reception block 10 based on the destination address and the port number of the IP packet. After decoding the RTP header, the UDP header, and the IP header, the packet receiving unit 100 outputs the payload of the packet (a code string according to ITU-T Recommendation G.729) to the decoding unit 200. Decoding section 200 is based on ITU-T Recommendation G. The 729 code string is decoded into the original audio sample string and stored in the jitter absorption buffer 300. The D / A converter 400 converts the voice into an analog signal at the sampling frequency calculated by a method described later, and the voice sample sequence is output as a sound wave by the receiver 500.
The above operation is the same as that of a conventional IP telephone when D / A conversion section 400 of receiving block 10 operates at a fixed sampling frequency.
[0014]
The packet receiving unit 100 receives a packet, and transmits a packet arrival trigger to the arrival interval calculation unit 700 when determining that the packet is a packet to be received. The arrival interval calculation unit 700 reads the current time from the clock 600 by the received packet arrival trigger, and calculates the difference from the previously recorded reception time of the previous packet arrival trigger. This difference is the packet arrival interval.
If the configuration is such that no jitter occurs before the packet is transmitted from the transmitting side, all packets should arrive at an interval of 10 milliseconds. Assuming that the calculated arrival interval is d milliseconds, the ratio R to the ideal arrival interval of 10 millimeters is given by the following equation (1).
R = d / 10
Is required.
Since the normal reproduction frequency is 8 kHz, the reproduction frequency F ′ [kHz] obtained by delaying the frequency of the D / A conversion by R is obtained as follows from equation (2).
F '= 8 / R
However, when the packet interval is observed for each packet, the operation of the D / A converter 400 becomes unstable when the packet interval is changed, and the operation of the D / A converter 400 becomes unstable. Use '. d 'is the average value of the past n arrival intervals.
When F ′ falls below the lower limit frequency Fmin, reproduction is performed at Fmin. In this case, the sound may be interrupted, but the reproduction below this will give a worse impression than the reproduction of the sound.
If the average arrival interval of the packets remains at 10 ms and only the jitter is large, d 'is about 10 ms, and the reproduction is performed at about 8 kHz.
If the average arrival interval of arriving packets deviates from 10 mm due to a decrease in the bandwidth, an increase in the processing time of the transmitting side, or a shift in the operating frequency of the transmitting side, reproduction at the corresponding frequency is performed. Breaks and skips are less likely to occur.
[0015]
As described above, according to the present embodiment, the reception interval of the audio packet data is calculated, and the reproduction speed of the audio data is controlled based on the reception interval, so that the data is not accumulated in the jitter absorption buffer. It is possible to prevent skipping of a sound or skipping of a sound due to the data using up the buffer area.
[0016]
As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to the embodiments, and may include a design change or the like without departing from the gist of the present invention.
[0017]
【The invention's effect】
As is apparent from the above description, according to the present invention, in order to control the operating frequency of the receiving side in consideration of the fluctuation of the operating frequency of the transmitting side and the receiving side and the fluctuation of the transmission delay, the voice packet data is jittered according to the present invention. In an IP telephone terminal provided with a D / A conversion unit that reads out from an absorption buffer and converts it into an analog signal, a reception interval is calculated from the reception time of the received voice packet data, and the conversion speed of the D / A conversion unit is calculated from the reception interval. Is calculated, and the D / A conversion unit is controlled so as to convert the latest voice packet data at the conversion speed. Sound skipping caused by overflowing from the area can be easily prevented.
In addition, there is no need for a mechanism using RTCP for clock synchronization between the transmitting side and the receiving side, which is simple.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of a receiving block 10 of an IP telephone 1 according to an embodiment of the present invention.
FIG. 2 is a block diagram showing a schematic configuration of an IP telephone 1 according to the embodiment.
FIG. 3 is a block diagram showing a schematic configuration of an arrival interval calculation unit 700 in the embodiment.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 ... Receiving block, 100 ... Packet receiving part, 200 ... Decoding part, 300 ... Jitter absorption buffer, 400 ... D / A conversion part, 500 ... Handset, 600 ... Clock, 700 ... Arrival interval calculation part, 1 ... IP telephone, 11 control block, 12 transmission block, 13 physical layer transmission / reception section, 121 transmission section, 122 A / D conversion section, 123 buffer, 124 encoding section, 125 packet transmission section

Claims (2)

An IP telephone terminal including a D / A conversion unit that reads out the audio packet data from a jitter absorption buffer that temporarily stores received audio packet data and converts the audio packet data into an analog audio signal.
Calculating means for acquiring the reception time of the voice packet data, calculating the reception time of the voice packet data from the reception time of the latest voice packet data and the reception time of the previous voice packet data,
Control means for calculating a conversion speed of the D / A conversion unit based on the reception interval, and controlling the D / A conversion unit to convert the latest voice packet data at the conversion speed. Characteristic IP telephone terminal. A data conversion method for an IP telephone terminal comprising a D / A converter for reading the voice packet data from a jitter absorption buffer for temporarily storing received voice packet data and converting the voice packet data into an analog voice signal.
A first step of obtaining a reception time of the audio packet data, calculating a reception interval of the audio packet data from a reception time of the latest audio packet data and a reception time of the previous audio packet data;
Calculating a conversion speed of the D / A conversion unit based on the reception interval, and controlling the D / A conversion unit to convert the latest voice packet data at the conversion speed. A data conversion method for an IP telephone terminal.

Images