Access and control system for enhanced audio capabilities in an integrated telephony/high speed data access device

[0003] This invention pertains generally to telephony, and more particularly to telephony over digital data networks.

[0004] Plain-old telephone service (POTS) has existed now for well over 100 years. During that time, people have grown used to the limited sound quality of POTS. At the outset of the telephone era, these limits on sound quality were due largely to the telephone's simple transducers and the transmission effects of long analog lines. As better transducers and compensation (gain and filtering) for transmission line effects became available, multiplexing transmission techniques also were introduced. These transmission techniques relied on the traditional limited audio bandwidth for voice communication, and thus the traditional limits on telephone sound quality remained.

[0005] As the public-switched telephone network (PSTN) evolved into a largely digitally-switched system between the 1960s and today, limited audio capabilities were maintained. The T1 digital carrier format, which is virtually ubiquitous within the PSTN today, relies on the appearance, at a PSTN analog-to-digital converter, of an analog telephone audio signal with a highest frequency of less than about 3500 Hz (sampled at 8 kHz and brick-wall filtered below 3500 Hz). Standard telephone handsets also cut off low-frequency transmissions to reduce 60 Hz power-line hum and other interference. These constraints limit the actual audio frequency range for a typical POTS phone conversation to 300-3400 Hz.

[0006] Packet-switched data networks now offer an alternative to PSTN telephony. Two computer users connected to a packet-switched data network can communicate data in any format that both understand. This includes packetized, digitized telephony in a variety of formats, including some offering stereo sound with a frequency range much larger than that offered by the PSTN and the standard POTS telephone. Due to cost and availability, current IP telephony equipment also uses traditional narrowband handsets.

[0007] Integrated telephony/high speed data access technology is now entering the marketplace. With this technology, a single wide-area-network (WAN) connection allows simultaneous packet data and packetized POTS telephone transmissions.

[0008] A packetized POTS transmission from a first integrated phone/data transceiver terminates either at a PSTN gateway (either digital or analog), or at a second integrated phone/data transceiver. The present invention recognizes that a packetized POTS transmission may be processed differently depending on how it terminates. For instance, if it terminates at a PSTN gateway, sound quality commensurate with the PSTN is all that makes sense. But if it terminates at a packetized POTS receiver, a much richer sound quality may be possible.

[0009] The present invention describes a telephone to digital data network interface. The interface comprises a telephone connection point (such as an RJ11 jack, electrical terminal pair, or a wire with an RJ11 plug) connected to a first selectable interface audio processing path for telephone-grade audio communication. The interface also comprises a second selectable interface audio processing path for wideband audio communication. An audio quality selector monitors the first processing path for the presence of switchover commands, and selects a desired audio processing path based on such a command when it is received from a telephone connected to the telephone connection point. Thus both standard telephone-grade and wideband telephony are available to the user and selectable from an attached telephone.

[0010] The telephone connection point preferably comprises an analog telephone line interface accepting a POTS phone. Preferably, the two audio processing paths use different digital audio decoders—one for wideband and one for telephone-grade audio. The second audio processing path preferably bypasses the telephone connection point, and, e.g., provides a connection for a headset, a stereo or television audio input, or a computer.

[0011] In another aspect, the invention provides a method of operating an analog telephone/digital data network interface. The method comprises the step of presenting a standard analog telephone line interface to an attached telephone set. But the interface detects a signal produced at the standard analog telephone line interface by the attached telephone set, and responds to the signal by reconfiguring audio processing between a telephone-grade processing path and a wideband processing path within the telephone/data network interface. Preferably the signaling is performed with DTMF codes, hookswitch flashes, voice prompts and commands, or a combination of these signaling techniques.

[0012] In yet another aspect, a cable modem transceiver is disclosed. This transceiver has connection points for remote data, local data, a telephone, and wideband audio. The transceiver further comprises a telephone to digital data network interface that has multiple switchable audio outputs, one of the switchable audio outputs connected to the telephone connection point and another of the switchable audio outputs connected to the wideband audio output connection point. The transceiver also comprises a bridge that routes data between the remote data connection point, the local data connection point, and the telephone to data network interface.

[0013] The invention may be best understood by reading the disclosure with reference to the drawing, wherein:

[0014] FIG. 1 illustrates a prior art integrated telephone/data access device;

[0015] FIGS. 2 and 3 show two possible network paths for calls directed through an integrated telephone/data access device;

[0016] FIG. 4 shows a telephone/data network interface according to the present invention;

[0017] FIGS. 5, 6, and 7 show alternate telephone/data network interface configurations according to the present invention; and

[0018] FIG. 8 shows a local configuration for an integrated telephone/data access device according to the present invention.

[0019] The following terms have the following meanings for purposes of this disclosure. Telephone-grade audio has an audio bandwidth with an upper frequency limit lower than 4500 Hz, and when in digital format, is digitally sampled at a 10 kHz or lower rate. Wideband audio has an audio bandwidth with an upper frequency limit greater than 6000 Hz, and when in digital format is digitally sampled at a 14 kHz or higher rate. A telephone is a device for two-way transmission of acoustic energy via electrical and/or electromagnetic signaling over a signaling network. A telephone contains at least two transducers, one that converts acoustic energy to electrical or electromagnetic energy, and one that performs the reverse of this process. A telephone connection point can either be analog or digital.

[0020] Integrated Telephony/Data Access Device Operation

[0021] FIG. 1 shows an integrated telephone/data access device 20 (e.g., a cable modem transceiver) comprising two main blocks, packet data bridge/router 36 and telephone/data network interface 34. Packet data bridge/router 36, operating in a well-understood manner, the details of which are not critical to the present invention, communicates over a data network connection 44, routing traffic appropriately between data network connection 44, local data network connection 32, and telephone/data network interface 34. Device 20 thus has the capability to perform data transfer while allowing simultaneous telephone access over data network connection 44.

[0022] Telephone/data network interface 34 looks like an analog telephone local loop to an attached phone, and like a packet data node to packet data bridge/router 36. Generally, interface 34 comprises an analog line interface 50 and a digital audio processor 70. Audio passing from analog line interface 50 to digital audio processor 70 passes through an analog-to-digital (A/D) converter 68. Audio passing in the opposite direction passes through a digital-to-analog (D/A) converter 66. A separate hookswitch signal 64 indicates to digital audio processor 70 whether an attached phone is on or off hook.

[0023] Analog line interface 50 emulates the analog line interface of the telephone company central office. In the exemplary circuit 50, power supply 54 provides talking battery and ringing voltage to RJ11 jack 52, and indicates on/off hookswitch conditions. Coils 58, 60, and 62 form a balanced transmitter. Driver 56 provides transmit audio drive current, and receiver 58 senses received audio.

[0024] Digital audio processor 70 handles the remaining phone processing tasks. Call control 76 handles call setup/takedown, and coordinates the operation of packetizer/depacketizer 78, decoder 74, encoder 72 and generates ring signal 46. Encoder 72 may compress digital audio received from A/D converter 68; encoder 72 may also perform voice activity detection (VAD), discrete-tone multi-frequency (DTMF, or “touch-tone”) detection, and echo cancellation. Decoder 74 may decompress digital audio received from the remote end of a phone connection. Decoder 74 may also provide tone generation (DTMF, dial-tone, ringback, busy signal) for line interface 50, and produce caller ID modulation signals. The details of these digital audio processor functions are all well understood in the industry and will not be repeated here.

[0025] Telephone/data network interface 34 allows a user to make and accept what appear to be conventional phone calls over a POTS phone 28 (see FIGS. 2 and 3). Calls actually pass through integrated telephone/data access device 20 and onto packet data network 22. The other end of the call may be similarly connected through an integrated telephone/data access device 40 onto packet data network 22, as shown in FIG. 3. But the other end of the call may also connect to PSTN 24, in which case a gateway 26 is needed to complete the connection, as shown in FIG. 2.

[0026] Data Network vs. PSTN Telephony Communication Paths

[0027] The human auditory system has a much wider range of operation than the PSTN. Normal human hearing perceives sound over a frequency range lying roughly between 20 Hz and 20,000 Hz, with the greatest sensitivity between about 60 Hz and 15,000 Hz. In the center of this frequency range, the human ear has a dynamic range of over 100 dB. By using both ears, a human can also perceive directional cues from intensity, phase, and timing differences in the sound received at each ear. But the PSTN, with its frequency range of 300 Hz to 3400 Hz, dynamic range of less than 50 dB, and monaural capability, limits audio communication to a level that, though intelligible, falls far below the limits of human perception.

[0028] Digital audio codecs (a matched audio encoder/decoder pair) span a large audio quality range. At the low end of audio quality, low bit-rate codecs such as those described in standards G.723.1, G.728, G.729, and G.729 Annex A, all promulgated by the International Telecommunications Union (ITU), generally produce sound quality inferior to the PSTN. ITU specification G.711 includes 8 kHz, 8 bits per sample PCM, generally following the digital standard used on the PSTN and thus offering comparable quality to the PSTN. Other codecs, such as those described in ITU specification G.722, offer 16 kHz or greater sample rates and thus offer the potential for wideband telephony. Compact-disk-quality stereo audio uses two 44.1 kHz sampled channels, each with a dynamic range of about 100 dB.

[0029] Despite the availability of wideband codecs, with a telephony audio data path as shown in FIG. 2, the transmission of audio with quality better than G.711 over packet data network 22 is unwarranted. POTS phones 28 and 30 are designed for PSTN bandwidths. Gateway 26 must also conform to PSTN bandwidths when it communicates with PSTN 24. These constraints limit conversation bandwidth, and thus a wideband codec used in such a communication path merely wastes data network bandwidth. Gateway 26 and telephone/data network interface 34 typically will transfer digital audio to each other using telephone-grade codecs, such as a G.711 codec set, or a low-bit-rate codec set such as G.729.

[0030] With a communication path as shown in FIG. 3, the transmission of audio with quality better than G.711 over packet data network 22 is still unwarranted. Although the PSTN is not part of the communication path, POTS phones 28 and 38 still use transducers with limited frequency capabilities. This is generally desirable, since these phones also are used for PSTN communications, and must meet PSTN specifications. But despite the limitations of POTS phones 28 and 38, an integrated telephone/data access device pair could be programmed to utilize any codec that fits hardware data bandwidth constraints and that both recognize, including G.722 or even a CD-quality codec.

[0031] A First Preferred Embodiment

[0032] The present invention recognizes that a POTS interface for an integrated telephone/data access device is desirable for many reasons. The POTS interface offers compatibility with existing phones and with the PSTN generally. It has been refined over more than a century of operation. It provides some privacy. It is familiar to everyone in a household from toddlers on up, even though the majority of these users may not understand or care whether their voice is being transmitted over a high-speed data network or a string.

[0033] At the same time, the present invention recognizes that the POTS phone is not revered for its audio quality, and that many POTS users would prefer better quality if it were available. In one aspect, the present invention thus seeks to combine a standard POTS interface with an option for enhanced audio. The resulting telephone to digital data network interface may be substituted for those of the prior art in an integrated telephone/digital data network access device, such as a cable modem.

[0034] In a first preferred embodiment depicted in FIG. 4, a telephone/data network interface 99 includes a POTS line interface 50 with, e.g., one or more attached RJ11 jacks 52. Telephone/data network interface 99 connects to a data network through digital audio processor 90. A/D and D/A converters 80 and 82 link line interface 50 with digital audio processor 90. In addition, an audio quality selector 86 (which may physically reside on digital audio processor 90) links line interface 50 with digital audio processor 90. Audio quality selector 86 is shown with a built-in DTMF detector 84, though a separate DTMF detector 84 (even one connected on the analog side of A/D 80) may be employed. Digital audio processor 90 has two decoders, a wideband decoder 98 (e.g., G.722-compliant) and a telephone-grade codec 96 (e.g., G.711-compliant). Call control 92 provides control lines to selectively enable one or the other of these decoders.

[0035] Audio quality selector 86 detects requests for audio quality reconfiguration that originate from a POTS phone attached to RJ11 jack 52. These requests may be entered from a conventional phone by hookswitch flash signaling, by depressing buttons on a numeric keypad to perform DTMF signaling, by voice command, or by a combination of these techniques. For example, audio quality selector 86 may be programmed to recognize two quick hookflashes, or a DTMF code such as “*77”, as a request for wideband audio. Audio quality selector 86 may also be configured to recognize a voice command such as “wide”, “stereo”, or “on TV” as a request for wideband audio. A voice command may be preceded by a signal such as a DTMF “*” to alert audio quality selector 86 to look for a voice command.

[0036] When audio quality selector 86 detects a request for audio quality reconfiguration that affects the codec used by digital signal processor 90, it relays the request to call control 92. Call control 92 contains an audio-grade renegotiator function. The renegotiator function examines whether the current codec provides the requested audio quality. If it does not, the renegotiator can attempt to renegotiate an appropriate codec with its opposite endpoint. Preferably, the renegotiator relies on a codec “preference list” that matches the requested audio quality.

[0037] Audio quality selector 86 may also “capture” signals designated for it. For instance, audio quality selector 86 may recognize all DTMF sequences input by the user, remove those directed to itself, and pass the remaining DTMF codes on to call control 92. Thus when an audio selection command is keyed or spoken by a user, the corresponding audio can be blanked instead of being interpreted by call control 92 and/or transmitted to a listener at the other end of the conversation.

[0038] Another desirable function for telephone/data network interface is an indication to a user that wideband audio is available from the other end of the conversation. At call connection, call control 92 performs codec negotiation with the transceiver at the opposite end of the data network connection using, e.g., the ITU H.323 standard or the SGCP (Simple Gateway Control Protocol). Call control 92 may negotiate for wideband capability first, or the opposite end may request wideband. In either case, call control 92 can implement a remote input bandwidth detector function—if wideband transmission is possible, the local user may be signaled, e.g., by an audible tone or a voice phrase such as “incoming wideband”. Alternately, a caller-ID signal producing a visual indication of wideband capability may be generated by call control 92.

[0039] The following pseudocode illustrates one possible algorithm, incorporating several of the above-described features, that call control 92 and audio quality selector 86 can use according to the present invention. This pseudocode operates if wideband capability is detected on the incoming digital audio channel. The pseudocode first indicates to the user that wideband is available. If the user has set a preference for automatic wideband switchover, this is performed. The code then goes into a loop for the duration of the call to process DTMF bandwidth switchover commands from the user.

[0040] Other Preferred Embodiments and Applications

[0041] Although the embodiment of FIG. 4 illustrates the basic control mechanism of the present invention, it will generally result in limited improvement in final audio quality, unless the attached POTS phone has been designed with an audio output transducer of well-above-average quality. The further preferred embodiments illustrate additional features of the invention, many of which allow greatly improved audio quality.

[0042] FIG. 5 illustrates a digital voice processor 90 that, in addition to the dual decoders 96 and 98, uses dual voice encoders 102 and 104. This system allows transmission, as well as reception, of dual-quality audio. Telephone-grade encoder 102 provides a default. Call control 92 can utilize wideband encoder 104 in one of several ways. It can switch to encoder 104 in tandem with switches to wideband decoder 96. It can allow the user to choose the encoder separate from the decoder. Or it can send wideband whenever the opposite end is equipped to receive it and/or requests it, with no user input required.

[0043] FIG. 6 shows a telephone/data network interface 118 with a switchable audio input and output, separate from the POTS phone connection point 52. A user of this system can use their POTS phone normally to initiate and receive phone calls. But by issuing a command over their POTS phone, the user can activate a switchable audio input 110 and/or audio output 112. The command may be the same command used to activate wideband encoders and decoders, or an independent command.

[0044] Audio quality selector 86 operates analog switches 106 and 108. Enabling analog switch 106 redirects incoming audio away from line interface 50 to an audio output 112 and 114. Enabling analog switch 108 redirects the local audio input away from line interface 50 to audio input 110 and 116. These switches could be ganged in a particular embodiment. Switch 106 could also have a position that supplies incoming audio to both line interface 50 and audio output 112. Switch 108 could also have a mixing position that combines audio received from line interface 50 with audio received from input 110.

[0045] When audio input is selected away from the POTS jack 52, POTS DTMF or voice commands cannot be processed with this particular embodiment. But commands may still be issued with the switchhook. For instance, a switchhook flash can be used to signal that audio is to be switched back to line interface 50. And an on-hook condition can still be used to end a call. When audio is selected away from the POTS phone, a separate source of commands (such as keyboard commands if audio is rerouted to a computer) may also be examined by audio quality selector 86.

[0046] An additional design consideration for a system according to the invention is digital sample rate. Common telephone-grade codecs operate at an 8 kHz digital sample rate, while common wideband codecs generally operate at 16 kHz or higher sample rates. A system according to the present invention must be compatible with both telephone-grade and wideband external sample rates.

[0047] One possible method of coping with different sample rates is to employ multi-rate conversion. FIG. 7 shows a telephone/data network interface 120 having multi-rate A/D and D/A converters 126 and 128. Audio quality selector 86 signals converters 126 and 128 as to the appropriate rate to use for conversion. This method may require a switchable anti-aliasing filter, shown incorporated with switch 122. This filter should be enabled to cut out high frequencies prior to A/D converter 126 when low-rate sampling is enabled, particularly if a wideband audio input is selected. It can be disabled, or a wideband anti-aliasing filter enabled instead, when high-rate sampling is enabled.

[0048] A second possible method of coping with different sampling rates uses single-rate converters that operate at the highest sample rate needed for wideband codecs used in the system. In this configuration, A/D and D/A converters are always run at this high rate. When the digital audio processor receives incoming remote audio at a lower sample rate, the audio is converted to the high rate prior to playout (e.g., by repeating samples or interspersing interpolating samples in the data) on the D/A. Outgoing audio is sampled at the high rate, regardless of the fidelity of the analog source. When a low-rate digital codec is enabled, the outgoing digital audio can be digitally low-pass filtered and then downsampled to the appropriate rate for presentation to the encoder.

[0049] The present invention can find application in an integrated telephony/wideband data device 130 as shown in FIG. 8. Data network connection 144 may, for a specific system, be a cable modem connection, an xDSL (a family of digital subscriber loops) modem connection, an ISDN (integrated services digital network) connection, or an IP baseband connection to a packet data network. Device 130 allows connection of multiple POTS phones 140 and 142. It also provides a digital data connection to a computer 132, an analog audio connection to a headset 134, and a connection to a set-top box 138 (which is in turn connected to a television 136). FIG. 8 is illustrative, and a particular device 130 may provide connection points for either more or fewer output devices than pictured.

[0050] Phone calls may be initiated and received over either phone 140 or 142 as described above. When a telephone user so desires, they can signal device 130 to enable wideband and/or to set one or more multiple position switches (e.g., like switches 106 and 108 of FIG. 6) to select among television 136, headset 134, and computer 132 for continuation of the call. A command for “on TV” could mean redirect audio output from the POTS handset to the TV audio, place the current TV audio programming in captioning mode, and negotiate for the wideband decoder. A command for “on headset” could mean redirect audio input and output from the POTS handset to a headset jack. And a command for “on PC” could mean activate an application on the computer and redirect audio input and output to the computer's microphone and speakers.

[0051] With respect to a computer, switching preferably would be done at a different point in device 130. Since computers generally have their own audio A/D and D/A converters, PCM audio streams may be sent in packets between device 130 and computer 132, bypassing the A/D and D/A converters of device 130. If the computer also has the capability to run the required audio codecs, the digital audio processor of device 130 may simply process control packets, and bridge audio packets between the network and the computer upon audio selection to the computer.

[0052] The user of device 130 may also select automatic preferences. For instance, incoming calls may be automatically directed to a location other than the POTS handset. Preferred configurations and associated commands may also be stored in device 130.

[0053] The present invention can also be incorporated into a system employing a digital phone, such as an IP phone or a digital PBX system. The operation of the present invention with such a system is similar to that described above, except that digital signaling such as a specialized “wideband” or “to computer” button on a digital phone can be programmed to initiate audio quality selection, and could still function even if audio were transferred elsewhere. In an IP phone, the phone bridges traffic to the selected wideband device. In a PBX system, the desk unit could bridge to a local wideband connection, or the central PBX switch could bridge, e.g., to the user's desktop computer. A system according to the invention may also use a phone that is integrated into a computer to issue audio quality selection commands to the telephone/data network interface.

[0054] After reading this disclosure, one of ordinary skill in the art will recognize that many advantageous modifications to the disclosed embodiments are enabled by the concepts taught herein. For example, although a packet data network has been used to illustrate the preferred embodiments, a time-division-multiplexed data network may also be used for data communication. The specific device features described herein may also be combined in a variety of permutations. The preferred embodiment of a digital audio processor and audio-quality selector is a software-configured digital signal processor—this includes partitioning of the described digital functions among several general purpose and/or special purpose processors, or utilization of a time partition on a processor also performing other tasks.

[0055] Other modifications to the disclosed embodiments will be obvious to those of ordinary skill in the art upon reading this disclosure, and are intended to fall within the scope of the invention as claimed.