KR100661511B1 - Jitter buffer device for a rtp of the voip system and controlling method therefore - Google Patents

Abstract

A jitter buffer device for RTP(Real Time Protocol) of a VOIP(Voice over IP) system and a controlling method thereof are provided to improve space efficiency of the jitter buffer by reducing memory usage space by constituting the buffer by separating memory spaces for processing sound data and video data. An input control part(1) separates video or sound VOIP data, and calculates storing position of the separated data and then writes the separated data at a corresponding storing position. A sound data buffer(2) stores sound data per connection in order to remove jitter according to a sound write control signal outputted from the input control part. A video data buffer(3) stores video data to remove jitter according to a video write control signal outputted from the input control part. An output control part(4) reads the video data or sound data from the video data buffer or the sound data buffer, and then transmits the video data or sound data to a destination.

Description

Jitter buffer device for a RTP of the VOIP system and controlling method therefore}

1 is an explanatory diagram illustrating input and output timing of a conventional jitter buffer.

Fig. 2 (a-b) is an explanatory diagram illustrating a conventional jitter buffer device for RTP.

Figure 3 (a-b) is an explanatory diagram for explaining the jitter buffer device for RTP of the VOIP system of the present invention.

4 is an explanatory diagram specifically illustrating the system of FIG. 3.

5 is a flowchart of the present invention.

<Detailed Description of Codes>

1: input control unit 2: voice data buffer

3: Image data buffer 4: Output control unit

5: data receiving unit 6: data separating unit

7: Voice light position calculation unit 8: Voice light control unit

9: image light position calculating unit 10: image light control unit

11: voice lead position calculator 12: voice lead controller

13: audio timer 14: video lead position calculator

15: video lead control unit 16: video timer

17: data multiplexer 18: data output unit

The present invention relates to a jitter buffer device for RTP and a control method of the VOIP system, and more particularly, to a RTP jitter buffer device and a control method of the VOIP system for implementing the jitter buffer to be divided into audio data and video data according to the traffic characteristics It is about.

In general, VoIP (voice over IP [Internet Protocol]) system is a term referring to IP telephony technology for a series of facilities for transmitting voice information using IP. In addition, the VoIP system means that voice information is transmitted in a digital form in discontinuous packets, rather than a circuit-based traditional protocol such as a public switched telephone network (PSTN). The main advantage of VoIP and Internet telephony technology is that it integrates telephone service by utilizing existing IP network so that telephone users can receive long distance and international telephone service in internet environment with only local call charge. In addition, the VoIP system uses Cisco and Vocaltek through VoIP forums to encourage the use of ITU-T H.323, a standard for transmitting voice and sound over IP over the public Internet or intranets within the enterprise. And major equipment manufacturers such as 3Com and NetSpeak. The VoIP Forum also encouraged directory service standards, enabling users to locate other users and enabling the use of touch phone signals for automatic telephone distribution and voicemail.

By the way, in the VOIP system as described above, the IP-based transmission and exchange method has become the mainstream of the current communication method. Therefore, in such a situation, the necessity of transmission using an IP network that has a wide range of data in the form of streams for real-time transmission, such as voice or video, rather than packet form, is increasing day by day. Therefore, the IETF enacted and recommended the RTP (real time protocol, RFC1889) protocol and is now widely used. Such RTP provides an end-to-end transport function suitable for applications that transmit real-time data such as voice, video, or mock data over multicast or unicast. However, the RTP does not deal with resource reservation, and in particular, does not provide functions such as guaranteeing QoS and preventing reversed transmission. Therefore, the meaning of transport can be said to be a standard established by focusing on the characteristics of real-time data. Therefore, it is necessary to supplement the function so that the user does not feel persistent or delayed, which is a jitter buffer. That is, as shown in FIG. 1, the jitter buffer has an irregular reception time in the short term even if the input data has a constant transmission rate in a long term. However, the output data passing through the jitter buffer After a slight delay, the output is maintained at a constant interval so that no empty space exists from the user's point of view.

Then, referring to the jitter buffer device of the conventional VOIP system with reference to FIG. 2, the data buffer unit 70 for storing the received voice and image data to remove the reception delay, and the data buffer unit 70 And an input control unit 71 for inputting and storing the received video and audio data, and an output control unit 72 for reading and transmitting the video and audio data stored in the data buffer unit 70 to a destination.

Here, the data buffer unit 70 may be formed in a structure for supporting a Gbe (Giga bit Ethernet) 1 port as shown in FIG.

Meanwhile, referring to the operation of the jitter buffer device of the conventional VOIP system, first, when the video or audio data is input from the external IP network or other terminal to the input control unit 71 of the VOIP system, the input control unit 71 The data is inserted into the data buffer unit 70 to eliminate time delay. The data stored in the data buffer unit 70 is read by the output control unit 72 and transferred to the corresponding destination, for example, an external IP network or other device unit.

However, the jitter buffer device of the conventional VOIP system is known that the size of the buffer to be implemented in the data buffer, that is, the jitter buffer should cover 120 m sec. The larger the buffer, the better, but the memory required for the implementation. The need for a lot and the delay is very high, but even if it is corrected, there is no big difference in quality from discarding because the receiving user experiences inconvenience due to the delay. Therefore, the conventional jitter buffer generally removes jitter within 120 msec. Therefore, the maximum size of the service packet is 40 bytes in the case of the video service, the characteristic of the video service is that a packet of 40 bytes unit is transmitted once every 5 msec to achieve an average speed of 64k bps (40x8x5 msec) in this state The number of buffers is 24 packets (120m / 5m). However, since the maximum number of connections per Gbe is 23000, 24 buffer units of 40 bytes are required for a total of 23000, that is, 40x8x24x23000 = 177Mbit to configure the jitter buffer. Therefore, the size of the jitter buffer memory as described above has a problem that the burden on the number of memory or the price is very large when implemented in SRAM, and when implemented in DRAM, the control is complicated and it is urgent to reduce the size of the memory. In other words, the jitter buffer device of the conventional VOIP system has a large amount of memory, and thus requires six to be used when using a current-produced maximum size SRAM (32Mbit). There is a problem.

Accordingly, the present invention has been invented to solve the above-mentioned problems, and since a buffer is formed by separating a memory space for processing voice data having a relatively low memory usage space and a video data having a relatively large memory usage space, a buffer space is used. It is an object of the present invention to provide a jitter buffer device for RTP of a VOIP system and a method of controlling the same, thereby significantly reducing the space efficiency of the jitter buffer.

It is still another object of the present invention to provide a jitter buffer device for RTP of a VOIP system and a method of controlling the same, which can significantly reduce the memory usage of the buffer and thus significantly reduce the configuration cost of the jitter buffer.

The present invention for achieving the above object is an input control unit for separating the video or audio VOIP data to be input, calculate the storage location of the separated data and then write to the corresponding storage location and output from the input control unit; A voice data buffer for storing voice data for each jitter according to the voice light control signal, and a video data buffer for storing the video data for jitter removal according to the image light control signal output from the input controller; A jitter buffer device for an RTP of a VOIP system and a control method thereof comprising an output control unit which reads video or audio data from a video data buffer or an audio data buffer and transfers the image or audio data to a set destination.

Another feature of the present invention is a data separation step of separating the currently input data into video data and audio data when data is received by the jitter buffer device of the VOIP system, and when the data currently separated after the data separation step is audio data. The audio data processing step of removing the jitter by inserting the separated audio data into the voice data buffer (34Mbit), then multiplexing with the video data to be transmitted to the corresponding destination, and if the data currently separated after the data separation step is the video data The present invention provides a control method of a jitter buffer device for an RTP of a VOIP system comprising a video data processing step of inserting separated video data into a video data buffer (41Mbit) to remove jitter, and then multiplexing the audio data and transmitting the same to a corresponding destination.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

The present invention provides an input control unit 1 for separating video or audio VOIP data to be input as shown in FIG. 3 (ab), calculating a storage location of the separated data, and writing to the corresponding storage location. According to the voice light control signal output from the input controller 1, the voice data buffer (2) for storing the audio data for each jitter to remove jitter, and according to the video light control signal output from the input controller (1) An image data buffer (3) for storing image data for jitter removal, and an output control unit (4) for transferring the image or audio data from the image data buffer (3) or the audio data buffer (2) to a set destination. Is made of.

As shown in FIG. 4, the input controller 1 separates the data received from the outside into data receiving unit 5 and data received from the data receiving unit 5 into audio data and video data. Using the data separating section 6, the voice write position calculator 7 for calculating an address which is a voice storage position from the voice data buffer 2, and the address received from the voice write position calculator 7 Voice light control unit 8 for storing the voice data received from the data separating unit 6 in the voice data buffer 2, and image light position calculation for calculating an address which is an image storage position from the image data buffer 3; An image data received from the data separator 6 in the image data buffer 3 by using the unit 9 and the address received from the image write position calculator 9; And a write control unit 10.

Further, the output controller 4 uses the voice lead position calculator 11 for calculating the data output position of the voice data buffer 2 and the address received from the voice lead position calculator 11. An audio lead control unit 12 for reading the position of the audio data stored in the data buffer 2, an audio timer 13 for supplying an audio lead trigger to the audio lead control unit 12, and the image data buffer 3 An image read position calculating section 14 for calculating a data output position of the image reading unit and an image read reading position of the image data stored in the image data buffer 3 using an address received from the image reading position calculating section 14; Image control unit 15, an image timer 16 for supplying an image read trigger to the image lead control unit 15, and an image and audio data output from the audio lead control unit 12 and the image lead control unit 15. And a data multiplexer 17 for multiplexing, and a data output unit 18 for outputting data received from the data multiplexer 17 to the specified destination.

In this case, the voice data buffer 2 is configured such that the required memory is 31x8x6x23000 = 34Mbit, so that the number of packets stored for 120msec is 6 (120msec / 20msec) and there are 23000 connections in total. On the other hand, the image data buffer 3 is configured by configuring the required memory as 40x8x24x5400 = 41Mbit, so that the number of packets stored for 120msec is 24 (120msec / 5msec) and there are 5400 connections in total.

Next, a control method applied to the above apparatus will be described.

As shown in FIG. 5, the method of the present invention proceeds to the data reception confirmation step S2 in the initial state S1 to check whether data is currently received by the jitter buffer device. At this time, if the data is not received by the jitter buffer device as a result of the data reception check step (S2), the current step ends and proceeds to the isle (IDLE) state. However, when the data is received by the jitter buffer device as a result of checking during the data reception confirmation step (S2), the data separation step (S3) is performed to separate the currently input data into video data and audio data. After the data separation step S3, the data checking step S4 is performed to confirm whether the currently separated data is image data. In this case, if the currently separated data is not video data as a result of the data checking step (S4), the process proceeds to the voice data processing step (S5) and puts the separated voice data into the voice data buffer (34Mbit). The jitter is removed and then multiplexed with the image data for transmission to the destination.

However, if the checked data during the data checking step (S4) indicates that the currently separated data is the image data, the process proceeds to the image data processing step (S6) to remove the jitter by putting the separated image data into the image data buffer (41Mbit) and then removing the voice. Multiplex the data and transfer it to the destination.

That is, in the present invention, when data is received from an external IP network or other device, the data input unit 5 inputs the input data to the data separator 6. Then, the data separator 6 separates the input data into voice data and image data, and transmits the input data to the voice light control unit 8 and the image light control unit 10, respectively. At this time, the voice light position calculating unit 7 receives the SN (Sequence Number) of the current packet from the voice light control unit 8 and information on the state in which the data of the current buffer is stored from the voice data buffer 2. Receives the write position of the data buffer 2 to calculate the transfer to the voice write control unit (8). In addition, the voice light control unit 8 stores the data received from the data separating unit 6 in the voice data buffer 2 using the address of the buffer received from the voice light position calculating unit 7. .

Meanwhile, for the separated image data, the image write position calculator 9 receives the SN (Sequence Number) of the current packet from the image write control unit 10 and receives data of the current buffer from the image data buffer 3. Receives the information on the stored state and calculates the write position of the data buffer 3 and transmits it to the image light control unit 10. In addition, the image light control unit 10 stores the data received from the data separating unit 6 in the image data buffer 3 using the address of the buffer received from the image light position calculating unit 9. .

On the other hand, when reading and processing the voice data, the voice read position calculator 11 receives the information from the voice data buffer 2 and how much data is stored in the current buffer to the voice read controller 12. Pass. The voice lead control unit 12 receives a trigger having a 20 msec interval from the voice timer 13 and generates a read signal at an address indicated by the voice lead position calculating unit 11 every 20 msec. Reads one packet per connection.

In addition, when the image data is read and processed, the image read position calculator 14 receives the information from the audio data buffer 3 and how much data is stored in the current buffer, to the image read controller 15. Pass. The image read control unit 15 receives a trigger having a 5 msec interval from the image timer 16 and generates a read signal at an address given by the image lead position calculating unit 14 every 20 msec to generate an image data buffer 3. Reads one packet per connection.

Meanwhile, the data read by the audio lead controller 12 and the image lead controller 15 are synthesized as one signal by the data multiplexer 17 and then transferred to the next step through the data output unit 18.

Therefore, a major difference between the present invention and the conventional structure is that the data buffer operates the buffer by separating the audio data and the video data according to the traffic characteristics. However, there are two main types of traffic serviced in WCDMA, one of which is AMR coded as 12.2K bps as voice data, one packet size is 31 bytes, and one packet is transmitted every 20 msec. can do. The other is video data, which is 64Kbps in speed, 40byte packet, and is transmitted one by one every 5msec. Therefore, such video data has a large occupied band per connection, so 5400 is the maximum connection in one GbE.

Therefore, the necessary memory is calculated by separating the audio and video data as follows.

Voice data memory: The number of packets stored during 120msec is 6 (120msec / 20msec) and there are 23000 connections, so the required memory is 31x8x6x23000 = 34Mbit.

Video data memory: The number of packets stored during 120msec is 24 (120msec / 5msec) and there are 5400 connections, so the required memory is 40x8x24x5400 = 41Mbit.

 Thus, the total size is 34M + 41M = 75Mbit. This is equivalent to 42% of the conventional 177M has the advantage of reducing the memory used more than half.

As described above, according to the present invention, the jitter buffer is implemented to be divided into audio data and video data according to traffic characteristics, so that the memory space for processing voice data having a relatively low memory footprint and video data having a relatively large memory footprint is processed. Since separate buffers can be used, the memory footprint can be considerably reduced, thus improving the space efficiency of the jitter buffer.

Further, according to the present invention, since the memory usage capacity of the buffer can be significantly reduced, the configuration cost of the jitter buffer is also significantly reduced.

Claims (8)

An input control unit which separates the input video or audio VOIP data, calculates a storage location of the separated data, and writes to the corresponding storage location, and removes jitter according to the audio light control signal output from the input control unit. An audio data buffer for storing the audio data for each connection, an image data buffer for storing the image data for jitter removal according to the image light control signal output from the input control unit, and an image or audio from the image data buffer or the audio data buffer An jitter buffer device for RTP in a VOIP system, characterized by comprising an output control unit for reading data and transferring the data to a set destination. The apparatus of claim 1, wherein the input control unit comprises: a data receiver for receiving data from an external source; a data separator for separating data received from the data receiver into audio data and video data; and a voice from the audio data buffer. A voice write position calculator for calculating an address which is a storage position, a voice write controller for storing voice data received from the data separator using an address received from the voice write position calculator, in a voice data buffer, and the image data buffer An image write position calculating unit for calculating an address which is an image storing position from the image writing unit, and an image write control unit storing the image data received from the data separating unit in the image data buffer using the address received from the image writing position calculating unit. Characterized Jitter buffer device for RTP in VOIP systems. The audio output apparatus of claim 1, wherein the output controller is configured to calculate a data output position of the voice data buffer, and a position of the voice data stored in the voice data buffer using an address received from the voice lead position calculator. A voice readout control unit for reading a voice readout, a voice timer for supplying a voice read trigger to the voice readout control unit, a video readout position calculating unit for calculating a data output position of the video data buffer, and an address received from the video readout position calculating unit. An image read control unit for reading the position of the image data stored in the image data buffer using the image readout, an image timer for supplying the image read trigger to the image read control unit, an image output from the audio lead control unit and the image read control unit, and A data multiplexer which multiplexes voice data, A data output device for a jitter buffer of RTP VOIP system characterized in that it comprises an output unit for the data received from the multiplexed data group to a set destination. The jitter buffer device for RTP of a VOIP system according to claim 1, wherein said voice data buffer comprises 34 Mbits of required memory. 5. The jitter buffer device for RTP of a VOIP system according to claim 4, wherein the voice data buffer has 6 (120 msec / 20 msec) packets stored for 120 msec and has a total of 23000 connections. The jitter buffer device for RTP of a VOIP system according to claim 1, wherein the video data buffer comprises 41 Mbits of required memory. 7. The jitter buffer device for RTP of a VOIP system according to claim 6, wherein the video data buffer has 24 (120 msec / 5 msec) packets stored for 120 msec and has a total of 5400 connections. When data is received by the jitter buffer device of the VOIP system, a data separation step of separating the currently input data into video data and audio data, and if the data currently separated after the data separation step is voice data, separate the separated voice data. The audio data processing step of removing the jitter in the audio data buffer (34Mbit) and then multiplexing the video data to transfer to the corresponding destination, and if the data currently separated after the data separation step is the video data, the separated video data The control method of the jitter buffer device for the RTP of the VOIP system, characterized in that the image data processing step of removing the jitter in the data buffer (41Mbit) and then multiplexed with the audio data to transfer to the destination.

Images