Classifications

• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04W—WIRELESS COMMUNICATION NETWORKS
  • H04W88/00—Devices specially adapted for wireless communication networks, e.g. terminals, base stations or access point devices
  • H04W88/18—Service support devices; Network management devices
  • H04W88/181—Transcoding devices; Rate adaptation devices
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
  • G10L19/16—Vocoder architecture
  • G10L19/173—Transcoding, i.e. converting between two coded representations avoiding cascaded coding-decoding
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
  • G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
  • H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L12/00—Data switching networks
  • H04L12/66—Arrangements for connecting between networks having differing types of switching systems, e.g. gateways

Definitions

• the present invention relates generally to the field of speech coding and, more particularly, to transcoding of speech in a packet network environment.
• VoP voice-over- packet
• VoP Voice data encoded according to one standard from a transmitting participant communicating in one network
• voice data encoded according to one standard from a transmitting participant communicating in one network has to be converted to the standard used by the receiving participant communicating under the guidelines of another network.
• a transmitting participant's speech may be encoded according to G.723.1 specifications while the receiving participant uses G.729.
• the bit-stream from the transmitting participant has to be converted from G.723.1 format to G.729 format.
• encoded data from the transmitting participant is decoded according to the coding method used by the transmitting participant.
• the decoded data is then re-encoded in accordance with the coding method used by the receiving participant.
• the data is transmitted to the receiving participant.
• Known transcoding schemes suffer numerous serious inadequacies.
• the decoding and re-encoding of the speech signal reduces the quality of the speech.
• the tandem operation of the post-filter common in low bit-rate speech decoders, can generate objectionable spectral distortion and degrade the speech quality significantly.
• a speech transcoder capable of transcoding a first bit-stream generated from a speech signal.
• the transcoder includes a decoder configured to receive the first bit- stream, which has been encoded based on a first coding scheme.
• the speech signal may have been encoded according to G.711, G.723.1, G726 or G.729, and may be parametric or non- parametric.
• the decoder extracts a plurality of first speech parameters from the first bit-stream, which may include, for example, parameters relating to spectral characteristics, energy, pitch and/or pitch gain of the speech signal.
• the decoder also decodes the first bit-stream according to the first coding scheme and generates a plurality of first speech samples.
• the decoder may include a post-filter element, which may be disabled to reduce system complexity and to improve the speech-quality of a speech signal generated by a subsequent re-encoding process.
• the plurality of first speech samples and the plurality of first speech parameters are then transmitted to a converter capable of converting the plurality first speech samples and plurality of first speech parameters to a plurality of second speech samples and a plurality of second speech parameters for use according to a second coding scheme.
• the second coding scheme may be G.711, G.723.1, G726 or G.729, for example, and may be parametric or non-parametric.
• the plurality of second speech samples and the plurality of second speech parameters are transmitted to an encoder.
• the encoder receives the plurality of second speech samples and plurality of speech parameters and generates a second bit-stream, the second bit-stream being encoded based on the second coding scheme.
• the encoder may include a noise suppressor element, which may be disabled to reduce system complexity and to improve the speech-quality of a speech signal. It is appreciated that by extracting the speech parameters from the first bit-stream, converting the speech parameters, and providing the converted speech parameters to the encoder avoids a re-evaluation of speech parameters during the encoding process, achieving many advantageous results, such as reduced system complexity and less delay.
• FIG. 1 illustrates a block diagram of a packet-based network in which various aspects of the present invention may be implemented
• FIG. 2 illustrates a block diagram of a transcoding system in accordance with one embodiment
• FIG. 3 illustrates a block diagram of a conference bridge utilizing a transcoding system in accordance with one embodiment
• FIG. 4 illustrates a block diagram of a component of a conference bridge utilizing a transcoding system in accordance with one embodiment
• FIG. 5 illustrates an exemplary flow diagram of a transcoding method utilizing the transcoding system of FIG. 2.
• the present invention may be described herein in terms of functional block components and various processing steps. It should be appreciated that such functional blocks may be realized by any number of hardware components and/or software components configured to perform the specified functions.
• the present invention may employ various integrated circuit components, e.g., memory elements, digital signal processing elements, logic elements, and the like, which may carry out a variety of functions under the control of one or more microprocessors or other control devices.
• the present invention may employ any number of conventional techniques for data transmission, signaling, signal processing and conditioning, tone generation and detection and the like. Such general techniques that may be known to those skilled in the art are not described in detail herein.
• FIG. 1 depicts an exemplary communication environment 100 that is capable of supporting the transmission of packetized voice information.
• a packet network 102 e.g., a network conforming to the Internet Protocol ("IP") , may support Internet telephony applications that enable a number of participants to conduct voice calls in accordance with conventional voice-over-packet techniques.
• packet network 102 may communicate with conventional telephone networks, local area networks, wide area networks, public branch exchanges, and/or home networks in a manner that enables participation by users that may have different communication devices and different communication service providers.
• Participant 1 and Participant 2 communicate with packet network 102 (either directly or indirectly) via the transmission of packets that contain voice data.
• Participant 3 communicates with packet network 102 via a gateway 104
• Participant 4 communicates with packet network 102 via a gateway 106.
• a gateway is a functional element that converts voice data into packet data.
• a gateway may be considered to be a conversion element that converts conventional voice information into a packetized form that can be transmitted over a packet network.
• a gateway may be implemented in a central office, in a peripheral device (such as a telephone), in a local switch (e.g., one associated with a public branch exchange), or the like. The functionality and operation of such gateways are well known to those skilled in the art and will therefore not be described in detail. It will be appreciated that the present invention can be implemented in conjunction with a variety of conventional gateway designs.
• a transcoder 108 may be included in packet network 102.
• Transcoder 108 may be implemented in a central office or maintained by an Internet service provider ("ISP").
• ISP Internet service provider
• the voice data from a number of packet-based participants, e.g., Participants 1 and 2 can be processed by transcoder 108 without having to perform the conversions normally performed by gateways.
• a transcoder 110 may be associated with or included in a gateway, e.g., gateway 104.
• transcoder 110 may be capable of receiving and processing voice-over-packet data and conventional voice signals.
• gateway 104 enables Participant 3, through transcoder 110, to communicate with packet network 102 and a participant coupled to packet network 102, e.g., Participant 1 or 2.
• a packet-based transcoder may be deployed in a telephony system to facilitate communication between participants using different standards or techniques of speech coding.
• a given packet-based voice channel may employ one of a number of different speech coding/compression standards.
• Various speech coding standards are generally known to those skilled in the art and may include, for example, G.711, G.726, G.728, G.729CA), G.723.1, Global System for Mobile Communication (“GSM”), selectable mode vocoder (“SMV”), and adaptive multi rate (“AMR”) coding, the specifications for which are hereby incorporated by reference.
• transcoder 108 or 110 may be capable of handling speech that has been encoded by various standards.
• transcoder should be capable of handling speech that has not been encoded.
• FIG. 2 illustrates exemplary communication system 200 for transcoding in accordance with one embodiment of the present invention.
• a first participant i.e., Participant 1
• a second participant i.e., Participant 2
• Transcoder 206 a first participant
• Participant 1 is coupled to transcoder 206 via channel 204
• Participant 2 is coupled to transcoder 206 via channel 216.
• voice data from Participant 1 may be encoded by encoder 202 and sent to transcoder 206 via channel 204.
• the voice data from participant 1 may need to be compressed and encoded by encoder 202 using a suitable coding standard.
• channel 204 may be an Internet-based packet network, in which case encoder 202 may use a suitable packet format to packetize the voice data.
• the output data from encoder 202 transmitted over channel 204 will include encoded digital data, in the form of a bit-stream, in accordance with one or more encoding standards, e.g., G.723.1 or G.729.
• channel 204 may function as a local link, coupling Participant 1 to transcoder 206, in which case encoder 202 may digitize the voice data from Participant 1 without encoding, and digitized data is transmitted over channel 204.
• the bit-stream from Participant 1 arriving at transcoder 206 via channel 204 is initially inputted into, and processed by decoder 208, which is configured to decode the bit-stream according to the coding method for the transmitting participant, i.e., Participant 1.
• decoder 208 would decode the bit-stream accordingly.
• post-filter element of decoder 208 may be disabled or its capabilities reduced to minimize the degradation frequently found with conventional decoding algorithms utilizing post-filtering.
• decoder 208 is also configured to extract certain speech parameters from the bit-stream.
• the speech parameters which are also referred to as "side information" in the present application, may include, for example, the energy, the spectral characteristics, the pitch and pitch gain of the speech signal.
• the speech parameters are transmitted by decoder 208 to converter 212.
• the speech samples and speech parameters inputted into converter 212 are suitably processed and converted for eventual encoding by an encoder according to the standard suitable for the receiving participant.
• the conversion performed by converter 212 may be based on the speech samples and/or at least one of the parameters, for example, received from decoder 208.
• the speech samples may be modified into a format suitable for re-encoding by encoder 214. For example, in instances where Participants 1 and 2 are using coding standards having different frame structures, converter 212 may resize the frames, to provide the speech samples according to a proper frame size for use by encoder 214.
• the speech information comprising the converted speech samples and speech parameters, is transmitted to encoder 214.
• decoder 208 may only provide the speech samples to converter 212 and no speech parameters (or side information).
• converter 212 receives the speech samples from decoder 208 and converts the speech samples to provide them according to a proper frame size for use by encoder 214.
• Encoder 214 is configured to encode the speech information according to the standard used by the receiving participant, i.e., Participant 2 in the present example. Thus, if Participant 2 utilizes a selectable mode vocoder ("SMV"), for example, then encoder 214 would encode the bit-stream according to the SMV standard.
• encoder 214 can be configured to encode speech information using the speech parameters extracted by decoder 208 and processed by converter 212. In this manner, parameters such as the energy, the spectral characteristics, the pitch and pitch gain of the speech signal, which are conventionally needed to re-encode the speech information by encoder 214, do not have to be re-extracted from the speech samples by encoder 214.
• encoder 214 does not have to perform such parameter estimation tasks as spectral analysis, pitch analysis, and the like, or encoder 214 may only have to perform lower complexity parameter estimation tasks.
• the transcoding scheme of various embodiments of the present invention substantially reduces the processing power, minimizes delay, and reduces overall system complexity when compared to conventional transcoding schemes.
• the noise suppression capability of encoder 214 may be disabled in order to further reduce the system's complexity.
• the speech parameters are extracted during the initial decoding step for use during the re-encoding step, degradation of the signal resulting from, for example, spectral and pitch re-evaluation is avoided.
• the bit-stream is transmitted to the receiving participant, i.e., Participant 2, via channel 216 in the format suitable for use by decoder 218, which then decodes the bit-stream.
• exemplary communication system 300 is used to illustrate a conference bridge using transcoding techniques of the present invention, in accordance with one embodiment. More particularly, communication system 300 shows how the present invention can be used to transcode and mix speech signals from two or more transmitting participants to a receiving participant, where each transmitting participant may be using a different coding scheme from the other.
• Participants 1, 2 and 3 are coupled to conference bridge 306 via channels 304, 316 and 322, respectively. It is appreciated that in the present example, Participants 1 and 3 are both communicating with Participant 2 at the same time.
• speech from participant 1 is encoded by encoder 302 into a format suitable for transmission over channel 304 to decoder 308.
• encoder 320 encodes speech from Participant 3 into a format suitable for transmission over channel 322 to decoder 324.
• Both decoders 308 and 324 can be configured to decode the incoming bit-streams, such as those coming from Participants 1 and 3, according to the coding schemes used by the transmitting participants and to generate speech samples from the bit-streams. Decoders 308 and 324 may also extract speech parameters, from the bit-stream, or generate the speech parameters if the speech was originally encoded according to a non-parametric standard.
• Converter/mixer 312 can be configured to convert, combine and mix the inputted speech samples and the speech parameters to generate a single speech information suitable for encoding according to the coding scheme used by the receiving participant, i.e., Participant 2.
• converter/mixer 312 may need to take into account frame size and other factors in order to generate a bit-stream suitable for encoding by the receiving participant.
• G.723.1 uses a frame size of 30 ms
• G.729 uses a frame size of 10 ms.
• a common frame structure may be established to enable effective mixing of the speech samples from decoders 308 and 324. For example, if at least one of the input channels is encoded using G.723.1, then a 30 ms frame may be established. Alternatively, a frame size equal to the least common multiple might be used.
• a 60 ms frame may be established. Once a frame size is determined, the speech samples and the speech parameters can be properly interpolated and aligned during mixing.
• converter/mixer 312 Once converter/mixer 312 has converted and mixed the signal from decoder 308 with the signal from decoder 324 to generate a combined bit-stream, the bit-stream is transmitted to encoder 314.
• Converter/mixer 312 can also provide encoder 314 with the speech parameters extracted from the inputted speech signals.
• Encoder 314 can be configured to re-encode the bit-stream according to the same coding standard used by Participant 2. For example, if Participant 2 uses G.726, then encoder 314 would re-encode the speech information according to G.726.
• Encoder 314 may use the parameters extracted by decoders 308 and 324 in order to re-encode the speech information, thus bypassing the need for spectral and pitch re-evaluation during the re-encoding process. In this manner, the complexity, processing demands, and time delay associated with such re-evaluation steps are avoided.
• the speech signal is transmitted via channel 316 to Participant 2, where decoder 318 decodes the signal.
• exemplary communication system 400 is used to illustrate a component of a conference bridge using transcoding techniques of the present invention, in accordance with one embodiment. More particularly, communication system 400 shows how the present invention provides an effective means for transcoding inputted speech signals having been encoded according to a non-parametric coding standard, such as G.711, G.726, and G.728, for example. As shown in FIG. 4, communication system 400 includes channel 404, conference bridge 406 and channel 416. It is appreciated that channels 404 and 416 are respectively equivalent to channels 204 and 216 of communication system 200 illustrated in FIG. 2.
• a speech signal transmitted to conference bridge 406 via channel 404 is decoded by decoder 408 to generate speech samples from the incoming bit-stream.
• Decoder 408 may also extract speech parameters from the bit-stream to generate the speech parameters in instances where the speech was encoded originally using a parametric standard, such as G.729 or G723.1.
• a parametric standard such as G.729 or G723.1.
• non-parametric speech coding standards for example G.711, G.726 and G.728, typically do not quantize various speech-related parameters, such as the signal pitch and spectrum. As a result, these parameters may not be extracted by decoder 408 directly from the bit-stream during the decoding process. In such instances, as shown in FIG.
• parameter extraction module 410 which extracts the desired speech-related parameters (or the side information) for subsequent use by encoder 414, as described below.
• parameter extraction module 410 can be configured to extract data regarding the signal energy, spectral characteristics, pitch and pitch gain, and the like, and to provide such parameters to converter/mixer 412.
• converter/mixer 412 receives speech samples and the speech parameters (or side information) 420 from other decoding devices (not shown).
• Converter/mixer 412 can be configured to combine and mix the speech samples and the speech parameters from decoder 408 and parameter extraction module 410 with speech samples and speech parameters 420 into a combined bit-stream suitable for use by encoder 414 in the re-encoding process. For example, in order to combine and mix the signals, converter/mixer may resize the frames of the speech samples in order to establish a common frame structure suitable for encoder 414.
• Converter/mixer 412 can also provide encoder 414 with the speech parameters (or side information) for use in re-encoding the bit-stream.
• the combined speech samples and extracted parameters provided by converter/mixer 412 can be used by encoder 414 to re-encode the speech signal according to the coding standard used by the receiving participant (not shown). Accordingly, by using the speech parameters (or side information) provided by converter/mixer 412, encoder 414 bypasses the need for spectral and pitch re-evaluation during the re-encoding process. In this manner, the complexity, processing demands, and time delay associated with such re-evaluation steps are avoided.
• the encoded signal is transmitted to the receiving participant via channel 416.
• FIG. 5, illustrates exemplary transcoding method 500 in accordance with one embodiment. It is appreciated that transcoding method 500 can be performed by a transcoder such as transcoder 206 in FIG. 1, for example. As shown, transcoding method 500 begins at step 510 and continues to step 512 where the bit-stream from a first participant is received.
• a parameter set is extracted from the bit-stream.
• the parameter set may include the signal energy, spectral characteristics, pitch and pitch gain, and the like.
• the bit-stream is decoded according to the coding scheme used by the first participant and speech samples are generated.
• the received bit-stream may be encoded according to G.723.1, in which case the bit-stream is decoded at step 516 according to G.723.1
• transcoding method 500 proceeds to step 518 where the speech samples and parameter set are converted into a suitable form for re- encoding.
• the form to which the speech samples and parameter set are converted may depend on the particular coding scheme used by the receiving participant.
• the converted speech samples are re-encoded in accordance with the coding scheme used by the receiving participant, i.e., the second participant in the present example. As such, if the second participant in the present description uses G.729, for example, then the re-encoding performed at step 520 would be done according to G.729.
• the re-encoding performed at step 520 can utilize the parameter set extracted from the bit-stream at step 516.
• transcoding method 500 provides a number of advantages over conventional transcoding approaches, including lower processing needs, minimal delay, and a reduction in overall system complexity.

Landscapes

• Engineering & Computer Science (AREA)
• Signal Processing (AREA)
• Physics & Mathematics (AREA)
• Health & Medical Sciences (AREA)
• Audiology, Speech & Language Pathology (AREA)
• Human Computer Interaction (AREA)
• Computational Linguistics (AREA)
• Acoustics & Sound (AREA)
• Multimedia (AREA)
• Computer Networks & Wireless Communication (AREA)
• Theoretical Computer Science (AREA)
• Telephonic Communication Services (AREA)
• Compression, Expansion, Code Conversion, And Decoders (AREA)

Applications Claiming Priority (3)

Publications (2)

Family

ID=29548267

Family Applications (1)

Country Status (7)

Families Citing this family (6)

Citations (4)

Family Cites Families (3)

• 2003
  • 2003-02-26 WO PCT/US2003/006335 patent/WO2003098598A1/en not_active Application Discontinuation
  • 2003-02-26 AU AU2003217859A patent/AU2003217859A1/en not_active Abandoned
  • 2003-02-26 JP JP2004506009A patent/JP2005531017A/ja not_active Withdrawn
  • 2003-02-26 KR KR10-2004-7017694A patent/KR20040104701A/ko not_active Application Discontinuation
  • 2003-02-26 CN CNA03810962XA patent/CN1653515A/zh active Pending
  • 2003-02-26 EP EP03713828A patent/EP1504441A4/en not_active Withdrawn
• 2004
  • 2004-11-10 IL IL16514704A patent/IL165147A0/xx unknown

Patent Citations (4)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events