KR20080050960A - Method for matching measuring interval for speech quality of multi-codec having variable bandwidth and apparatus using the same - Google Patents

Abstract

A method and an apparatus for matching measurement interval of speech quality of variable band multi-codec are provided to match a measurement interval between a reference sound and a codec sound and reduce a measurement result distortion phenomenon. A measurement interval setting unit(110) sets an interval of speech quality desired to be measured by time intervals, and the first conversion unit(120) converts the measurement interval by sample units. An information synthesizing unit(130) synthesizes information about the measurement interval which has been converted by sample units and a digital original sound, and outputs the same.

Description

Method and matching measuring interval for speech quality of multi-codec having variable bandwidth and apparatus using the same

The present invention relates to a method and apparatus for matching voice quality measurement intervals when measuring variable band multicodec voice quality, and more particularly, to provide a real-time multimedia service by performing an interworking function between a packet network and an existing wired or wireless network. When the calling party and the called party use a variable band multicodec that provides different rates, the transmission rate of the variable band multicodec according to the network's forwarding capability, the network's propagation delay when delivering the network, packet loss, CPU utilization, natural sound, A feature that compares the same voice intervals in the speech quality measurement, which compares the natural sound with the codec output sound or the sound after network transmission in a speech quality measurement device that enables the comparison between the codec output sound, the codec output sound, and the sound after network transmission. It relates to an apparatus and a method for providing the same.

The present invention is derived from a study conducted as part of the IT new growth engine core technology project of the Ministry of Information and Communication and the Ministry of Information and Telecommunication Research and Development. [Task Management No.:2006-S-100-02, Title: Variable-band multicodec and control technology ].

Variable-band multicodec is a technology for converting natural sound into digitally converted codec data having multiple data rates. For example, narrow band (band from 300 to 3,400 Hz), wide band (band from 50 to 7,000 Hz) , Codec technology that can separate audio bands (bands up to 20-20,000Hz) and calculate the transmission rates of 8, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32bps in each band to be.

For example, in a VoIP voice telephony service of a packet network, the bandwidth provided by the network is variable and unpredictable. The 32bps transmission rate of variable band multicodec is the codec rate that yields the best sound quality, while the 8kbps rate is the case where the codec rate provides the lowest sound quality. If the network bandwidth is sufficient to send a high bandwidth, 32kbps transmission band is transmitted, and if the network situation changes due to the change of network bandwidth, the sound quality is bad by lowering the transmission rate such as 30kbps, 28kbps, .. Be sure to In a variable band multicodec, the sound quality is good when the transmission rate is high, but transmission loss and delay are high in the network, and the sound quality is low when the transmission rate is low, but the transmission loss and delay in the network is low.

The voice quality of the variable band multicodec can be measured end-to-end so that accurate transmission can be controlled by accurate situation determination. Such accurate data rate control requires accurate voice quality measurement. The voice quality measurement is performed by comparing a file recording natural sound (recording file_1) with a file recording codec output sound (recording file_3). In this process, when the voice sections included in the two files are different, accurate voice quality measurement cannot be performed.

Meanwhile, a signal protocol conversion technique for call setup is applied to apply a variable band multicodec. These signal protocol conversion technologies are disclosed in IETF's RFC 3261 SIP, RFC 3264 Offer / Answer SDP, RFC 2833 RTP Payload for DTMP Digits, Telephony Tones and Telephony Signals, RFC 2327 SDP, RFC 3108 ATM SDP, RFC 1890 RTP Profile Payload type, etc. It is.

Conventional techniques for determining the end-to-end speech quality measurement method of the codec are as follows.

Apparatus and method for measuring call quality in a mobile communication system, which is Korean Laid-Open Publication No. 2003-0019839, are devices capable of measuring call quality in real time in a mobile communication system.

The method of automatically measuring the vocoder quality of the CDMA system of Korean Laid-Open Publication No. 2000-0025237 is an apparatus for automatically measuring the quality of the vocoder mounted in the control station in the CDMA system.

US Patent No. US 7,002,992 B1, titled Codec Selection to improve media communication, is a device for selecting a codec by a network parameter.

US Pat. No. 6,944,126B1, entitled Method for determining loss rate or a loss curve of a video source or multimedia traffic, is a device for adjusting the buffer size for a given rate.

US Pat. No. 6,512,746 B1 Method and apparatus fo measuring voice grade of service in an IP network is a non-intrusive measuring device to ensure a service level agreement.

In Korean Publication No. 10-2004-0011257 (2004.2.5), a code rate of MPEG is selected based on a buffer size. In addition, Korean Publication No. 10-2000-0072520 (2000.12.5) proposes a voice data priority transmission method. Republic of Korea Publication No. 10-2004-0049559 (2004.6.12) proposes a method for securing the resources of the Internet packet data based on the connection establishment protocol.

US Patent No. 6,014,694 (Jan. 11, 2000) proposes a method of controlling between a video server and a receiver. In addition, US Publication No. 2003/0119515 A1 (Jun. 26, 2003) proposes an automatic configuration method for a codec in a mobile communication CDMA base station. US Patent Publication No. 2004/0066763 A1 (Apr. 4, 2004) proposes a method of using QoS packets to improve forwarding capability in a wireless LAN. US Publication No. 2003/0189900 A1 (Oct. 9, 2003) proposes an AMR codec application method for mobile communication using frequency line rate and error rate of an air interface. U.S. Publication No. 2005/0201414 A1 (Sep. 15, 2005) proposes a method for optimizing IP overhead on Ethernet by sending multiple codec frames per packet in Internet telephony.

U.S. Publication Nos. 2004/0002856 A1 (Jan. 1, 2004) and 2005/0055203 A1 (Mar. 10, 2005) present a multi-rate frequency domain speech codec system. do. US Patent Publication No. 2004/0185786 A1 (Spt. 23, 2004) proposes a method using a Predictor output frame structure in an advanced mobile communication device. US Patent Publication No. 2004/0213257 A1 (Oct. 28, 2004) proposes a method for negotiating the establishment of an appropriate channel for a predetermined codec transmission. US Patent Publication No. 2004/0160979 A1 (Aug. 19, 2004) proposes a method of controlling an AMR codec using source coding and channel coding. U.S. Publication No. 2004 / 5,657,420 A1 (Aug. 12, 1997) proposes a codec scheme of Qualcomm in the United States for variable rate vocoder.

This conventional technique has a problem in that it is not possible to find the cause of the distortion of the voice quality during the selection of the end-to-end voice quality measurement target of the present invention. Therefore, by using the voice quality measurement algorithm between the natural sound, the variable band multi-codec conversion and the collection of files received after the network transmission according to the change of the state of the network band, the cause of the distortion of the voice quality is found. There is a need for a method and apparatus.

The technical problem to be achieved by the present invention is to match a real-time voice quality measurement interval between a natural original sound (hereinafter referred to as "digital original sound") and a real-time multimedia service sound (hereinafter referred to as "digital inverse conversion sound") in a real-time multimedia service. The present invention provides a voice quality measurement interval matching device and a method for achieving accurate voice quality measurement.

In order to achieve the above technical problem, a transmission apparatus for matching a variable band multi-codec voice quality measurement interval according to the present invention includes a measurement interval setting unit for setting a time interval for measuring voice quality; A first converter converting the measurement interval into a sample unit; And an information synthesizing unit for synthesizing and outputting the information about the measurement interval converted to the sample unit and the digital original sound.

In order to achieve the above technical problem, a receiving apparatus for matching a variable band multicodec voice quality measurement interval according to the present invention is a device for measuring voice quality of a variable band multicodec by receiving information about a measurement interval and digital original sound. An information extraction unit for extracting information about the measurement interval and digital original sound from the received data; A second converting unit converting the extracted information into byte units; And a voice extractor extracting a sound quality interval measurement sound based on the byte unit information.

According to an aspect of the present invention, there is provided an apparatus for matching a variable band multi-codec voice quality measurement interval according to the present invention, comprising: a measurement information transmitter for synthesizing information about a voice quality measurement interval and a digitized original sound into one data; And a measurement information receiver which receives the data and separates the information and the digitized original sound to generate voice data for a measurement section based on the information.

In order to achieve the above technical problem, a method for matching a variable band multi-codec voice quality measurement interval according to the present invention is a method of measuring voice quality by connecting a voice quality measurement transmitter and a voice quality measurement receiver to a network. Transmitting, by the voice quality measurement transmission apparatus, information about the voice quality measurement interval and the digitized original sound as one data; And generating, by the voice quality measurement receiving apparatus, the data and separating the information and the digitized original sound to generate voice data for a measurement section based on the information.

As described above, according to the method and apparatus for matching variable-band multicodec voice quality measurement intervals according to the present invention, the variable-band multicodec is used as a reference for measuring voice-to-end voice quality in real-time Internet multimedia services. It is possible to provide a method of matching measurement intervals between the reference sound and the codec sound to be measured, and to reduce the distortion of the measurement result due to the measurement interval mismatch.

In addition, it is possible to store the data of the voice quality measurement target with accurate start and end points in each of the digital original sound and the digital reverse conversion sound for accurate quality measurement, and calculate the accurate voice quality measurement value.

In addition, a variable-band multi codec with a variable network transmission bit rate is delivered to the network to provide a method for measuring voice quality and to determine the cause of the voice quality distortion as a result of the measurement. Can be used.

The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, whereby those skilled in the art may easily implement the technical idea of the present invention. There will be. In addition, in describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1A is a block diagram illustrating a configuration of a transmitting device for matching a variable band multicodec voice quality measurement interval according to the present invention, and FIG. 1B is a configuration of a receiving device for matching a variable band multicodec voice quality measurement interval according to the present invention. It is a block diagram showing. 2 is a block diagram showing the configuration of an apparatus for matching a variable band multicodec voice quality measurement interval according to the present invention. 3 is a flowchart showing a process of a method for matching a variable band multicodec voice quality measurement interval according to the present invention.

In the following embodiment, the call processing flow between the calling party and the called party will be described taking the case of processing the IETF standard as an example. At this time, the call flow processed by the IETF standard specification from the called party to the originating party follows only the standard specification and describes only the invented contents not included in the standard specification.

Reference is first made to FIG. 1A. The measurement information transmitting device includes a measurement section input unit 110, a first converter 120, and an information synthesizer 130, receives a desired voice quality measurement section from a user in units of time, and converts the sample into sample units. It has a function of synthesizing the measurement section information into one data together with the digitized original sound and transmitting it to the measurement information receiving apparatus as shown in FIG. 1B.

The measurement interval setting unit 110 receives the desired voice quality measurement interval in units of time from the user and then sends this information to the first conversion unit 120 (step S301).

The first converter 120 converts the voice quality measurement section of the time unit received by the measurement section setting unit 110 into sample units. The voice quality measurement interval information of the converted sample unit is transmitted to the information synthesizing unit 130 (step S303).

The information synthesizing unit 130 synthesizes the voice quality measurement interval information in the sample unit received from the first converter 120 and the digital original sound 105 into one data. The synthesized data is finally output synthesized data and is transmitted to the receiving device through the network (S305).

The digital original sound is the original sound that is a standard for measuring voice quality obtained by converting a micro-input analog voice into digital.

The measurement information receiver shown in FIG. 1B includes an information extractor 140 for extracting information about the measurement interval and the digital original sound from the received synthesized data, and a second converter 150 for converting the extracted information into bytes. And a voice extractor 160 for extracting a sound quality interval measurement sound based on the byte unit information, and the voice extractor 160 encodes and decodes the digital original sound through a codec. A first voice extractor 161 for extracting an inverse transform sound for measurement by an interval based on the information from an inverse transform sound and a second voice extractor 163 for extracting an original sound for an interval based on the information from the digital original sound; Is done. And it is preferable to include a quality measuring unit 170 for measuring the quality by comparing the measurement reverse conversion sound and the measurement original sound.

The information extractor 140 extracts voice quality measurement interval information in a sample unit from the synthesized data received from the network. The separated voice quality measurement interval information is sent to the second converter 150, and the digital original sound is sent to the second voice extractor 163 (step S307).

The second converter 150 converts the voice quality measurement interval information in the sample unit into the byte unit of the voice data. The converted voice quality measurement interval information in units of bytes is sent to the first voice extractor 161 and the second voice extractor 163, respectively (step S309).

The first speech extractor 161 receives the digital inverse transform sound encoded and decoded through the codec and outputs the inverse transform sound for measurement using the speech quality measurement section information received from the second transform unit 150. The measurement inverse transform sound extracts corresponding interval data from the first inverse transform sound data using interval information (bytes A to B bytes, B is larger than A) in byte units. In other words, the first byte of the first inverse transform data starts with one byte and uses data from A bytes to B bytes as the inverse measurement sound for measurement.

The second voice extractor 163 outputs the original sound for measurement using the voice quality measurement interval information received from the second converter 150 from the digital original sound received from the second measurement interval information extractor 108. The original sound for measurement also uses the same method as that of generating the inverse transform sound for measurement (step S311).

The digital inverse transform sound is speech data obtained by processing digital natural sound encoded and decoded through a codec.

The measurement inverse transform sound is voice data extracted only by the voice quality measurement interval from the digital inverse transform sound.

The original sound for measurement is voice data extracted only from the digital original sound by the voice quality measurement section.

FIG. 2 is a block diagram showing an example of a configuration in which the measurement information transmitter 205 and the measurement information receiver 211 of FIG. 1A are applied to a voice quality measurement apparatus. The natural sound detector 201 and the voice converter 202 and measurement information are shown in FIG. Transmitter 205, network transmitter 206 and receiver 208, IP network 207, voice inverse converter 209, measurement information receiver 211, voice quality meter 213, natural sound reproducer 212 ) And the like.

The natural sound detector 201 is a device for digitally converting human or natural sounds to produce the digital original sound 204, which is transmitted to the voice converter 202 and the measurement information transmitter 205. .

The voice converter 202 encodes the digital original sound 204 converted by the natural sound detector 201 to generate the digital conversion sound 203. The generated digital tones 203 are sent to the network transmitter 206 to send through the network to the quality measurement receiver.

The measurement information transmitting apparatus 205 receives a voice quality measurement section in units of time through a user input, converts it into a sample unit, and synthesizes the converted voice quality measurement section information and the digital original sound 204 into one data. This shows the measurement information transmitter shown in Fig. 1A.

The network transmitter 206 receives the digital conversion sound 203 converted by the voice converter 202 and the synthesized data synthesized by the measurement information transmitter 205 and transmits the synthesized data to the network. Network protocols TCP / IP and UDP / IP are used.

Data transmitted through the IP network 107 is collected through the network receiver 208 of the quality measurement receiver and the digital original sound 204 and voice quality measurement information using TCP are stored back into the synthesis data 214, The digital conversion sound 203 using UDP is stored as the digital inversion sound 210 through the voice inverse converter 209.

The synthesized data 214 and the digital inverse transform sound 210 collected at the receiver are extracted from the measurement information receiving apparatus 211, and the voices of the sections necessary for the measurement are transferred to the voice quality measuring instrument 213 to perform quality measurement. The natural sound player 212 receives the digital inverse transform sound 210 to reproduce the sound.

The network receiver 208 extracts the desired data according to the RTP / UDP and TCP standards.

The voice inverse converter 209 performs a decode function of inversely converting voice data received through a network into data that can be reproduced by a natural sound player.

The measurement information receiving apparatus 211 receives the synthesized data including the voice quality measurement interval information and the digital original sound from the network receiver 208 and separates them, respectively, and uses the separated voice quality measurement interval information, the voice inverse converter 209. Only the interval data necessary for measuring voice quality is extracted from the digital inverse transform sound 210 received from the digital original sound. This shows the measurement information receiving apparatus shown in FIG.

The voice quality measurer 213 receives the digital original sound and the inverse conversion sound for measurement from the measurement information receiving apparatus 211 and performs voice quality measurement using various algorithms.

The natural sound player 212 reproduces the digital inverse transform sound 210.

The method for matching a variable band multicodec voice quality measurement interval according to the present invention can also be embodied as computer readable codes on a computer readable recording medium. Computer-readable recording media include all kinds of recording devices that store data that can be read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disks, optical data storage devices, and the like, which are also implemented in the form of carrier waves (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing the present invention can be easily inferred by programmers in the art to which the present invention belongs.

Such a method and apparatus of the present invention have been described with reference to the embodiments shown in the drawings for clarity, but these are merely exemplary, and various modifications and equivalent other embodiments are possible to those skilled in the art. Will understand. Therefore, the true technical protection scope of the present invention will be defined by the appended claims.

FIG. 1A is a block diagram illustrating a configuration of a transmitting apparatus for matching a variable band multicodec voice quality measurement interval according to the present invention.

1B is a block diagram illustrating a configuration of a receiving apparatus for matching a variable band multicodec voice quality measurement interval according to the present invention.

2 is a block diagram showing the configuration of an apparatus for matching a variable band multicodec voice quality measurement interval according to the present invention.

3 is a flowchart illustrating a method of matching a variable band multicodec voice quality measurement interval according to the present invention.

Claims (14)

A measurement section setting unit for setting a section to be measured for voice quality in units of time; A first converter converting the measurement interval into a sample unit; And And an information synthesizing unit for synthesizing the information on the measurement section converted in the sample unit and the digital original sound and outputting the synthesized audio signal. An apparatus for measuring voice quality of a variable band multicodec by receiving information about a measurement interval and digital original sound, An information extraction unit for extracting information about the measurement interval and digital original sound from the received data; A second converting unit converting the extracted information into byte units; And And a voice extractor extracting a sound quality interval measurement sound based on the byte unit information. The method of claim 2, wherein the voice extraction unit A first speech extracting unit for extracting the inverse transform sound for measurement by the section based on the information from the digital inverse sound converted through the codec and encoded and decoded; And And a second voice extractor for extracting the original sound for the measurement as much as the interval based on the information from the digital original sound. The method of claim 3, wherein the voice extraction unit And a quality measuring unit configured to perform voice quality measurement on the measurement inverse conversion sound and the original sound for measuring. A measurement information transmitter for synthesizing and transmitting the information about the voice quality measurement interval and the digitized original sound into one data; And And a measurement information receiver which receives the data and separates the information and the digitized original sound to generate voice data for a measurement section based on the information. . The method of claim 5, wherein the measurement information transmitting unit A measurement section setting unit for setting a section to be measured for voice quality in units of time; A first converter converting the measurement interval into a sample unit; And And an information synthesizing unit for synthesizing the information on the measurement interval converted into the sample unit and digital original sound and outputting the synthesized sound. The method of claim 5, wherein the measurement information receiving unit An information extraction unit for extracting information about the measurement interval and the digital original sound from the received data; A second converting unit converting the extracted information into byte units; And And a voice extractor for extracting a sound quality interval measurement sound based on the byte unit information. The method of claim 7, wherein the voice extraction unit A first speech extracting unit for extracting the inverse transform sound for measurement by the section based on the information from the digital inverse sound converted through the codec and encoded and decoded; And And a second voice extractor for extracting the original sound for the measurement as much as the interval based on the information from the digital original sound. In the voice quality measurement transmitter and the voice quality measurement receiver connected to the network to measure the voice quality, (a) synthesizing and transmitting the information on the voice quality measurement interval and the digitized original sound into one data by the voice quality measurement transmitting apparatus; And (b) the voice quality measurement receiving apparatus receiving the data, separating the information and the digitized original sound, and generating voice data for a measurement interval based on the information; variable band multicodec Method for matching speech quality measurement intervals. The method of claim 9, wherein step (a) (a1) setting a voice quality measurement interval in time units; (a2) converting the voice quality measurement interval in units of time into a sample unit which is a unit for converting an analog signal into a digital signal; And and (a3) synthesizing the voice quality measurement interval information converted into a sample unit and the digital original sound into one data unit to generate synthesized data. The method of claim 9, wherein step (b) (b1) extracting information about the measurement interval and the digital original sound from the received data; (b2) converting the extracted information into units of bytes; And (b3) extracting a sound quality interval measurement sound based on the byte unit information; and a variable band multi-codec voice quality measurement interval coincidence. The method of claim 11, wherein step (b3) (b31) extracting the inverse measurement sound for the measurement by the interval of the information from the digital inverse sound that has been encoded and decoded through the codec through the codec; And and (b32) extracting the original sound for the measurement by the interval based on the information from the digital original sound. The method of claim 12, And performing voice quality measurement on the measurement inverse converted sound and the original sound for measurement. 6. In the voice quality measurement transmitter and the voice quality measurement receiver connected to the network to measure the voice quality, (a) synthesizing and transmitting the information on the voice quality measurement interval and the digitized original sound into one data by the voice quality measurement transmitting apparatus; And (b) the voice quality measurement receiving apparatus receiving the data, separating the information and the digitized original sound, and generating voice data for a measurement interval based on the information; variable band multicodec A computer-readable recording medium that records a program that can be executed on a computer to match a voice quality measurement interval.

Images