WO1996032804A1

WO1996032804A1 - Device and method for voice conference communication

Info

Publication number: WO1996032804A1
Application number: PCT/SE1996/000438
Authority: WO
Inventors: Stig Stuns; John-Erik Eriksson; Hans EKSTRÖM; Gunnar EKSTRÖM
Original assignee: Konftel Ab
Priority date: 1995-04-07
Filing date: 1996-04-03
Publication date: 1996-10-17
Also published as: AU5410996A; SE9501681D0

Abstract

The invention relates to a voice communication device (20) and to a method of recording and transmitting voices and replaying the voices in full duplex. The invention relates more specifically, but not exclusively, to a conference telephone device. The invention also relates to effective cancelling of line echoes and prevents the occurrence of acoustic feedback. The invention also relates to an arrangement comprising a centrally disposed all-directional recording microphone (28) surrounded by several loudspeaker units (16) in its immediate vicinity. The invention also relates to a device provided with a remote keypad (70) for cordless dialing, e.g. via an IR pulse train (80), from a loudspeaker telephone or central unit (60) from any selected place in a conference room, for instance. According to one particular embodiment, the keypad is used for a loudspeaker telephone having a sound unit comprised of a centrally located all-directional recording microphone (28) and at least two loudspeaker units (60) disposed in the immediate vicinity of the microphone.

Description

DEVICE AND METHOD FOR VOICE CONFERENCE COMMUNICATION TBCEHICλL "FIELD

The present invention relates to a device for voice communi¬ cation and a method of recording and transmitting voices and playing-back the voices in full duplex, and a cordless remote keypad for said device. The invention specifically addresses a conference telephone device, with a cordless remote keypad for loudspeaker telephones. DESCRIPTION OP THE BACKGROUND AST

Those systems for conference telephony that include loud¬ speaker telephones and so-called full duplex, i.e. with two- way speech, commercially available at present are constructed around a central sound unit, which is preferably placed on a table with the conference members uniformly placed around the unit. In addition to the means required to set up and maintain telephone call connections, the known sound units also include a loudspeaker which is arranged centrally on the units and which has three microphones disposed symmetrically around the loudspeaker.

Because the microphones face in specific directions, the conference members must be seated roughly in an organized pattern around the sound unit. Should a conference member rise quickly from his/her position around the table, speech attenuation will occur in the known systems. The reason why the microphones are directed in this way is due to acoustic feedback in the system, which gives rise to noise disturbanc- es, such as intolerable singing or howling noise.

Another problem associated with known conference sound units is the line echo that occurs when several sound units are connected together. For instance, the sound units may be connected to different telephone systems, a feature that will become more prevalent in the present-day deregulated tele¬ phone industry. This means that the systems often give rise to different line echoes for connection to the telephone network, and hence signals will be reflected with signal atenuation as a result.

The situation is further complicated by the fact that ongoing conversations are echoed in each room in which a sound unit is placed.

In order to exemplify the aforesaid problems, there is given below an extract of the recommended use of a modern confer¬ ence telephone marketed by a large commercial company. The following recommendations are made under the heading "In Order to Obtain the Best Sound Quality":

- Place the sound unit in a quiet environment, preferably an environment furnished with furniture, curtains and other sound-absorbing materials.

Place the sound unit in the middle of the table.

Papers, documents and other objects are preferably placed slightly away from the sound unit.

Position the conference members roughly equidistantly from the sound unit.

Speak in a normal voice.

Turn towards the sound unit when speaking. - Do not move around the sound unit when speaking.

These recommendations illustrate the problems associated with high-quality conference telephony.

Those services that are available in modern telephone systems which include, e.g., telephone switching centres and exchang¬ es manufactured by Ericsson, e.g. in accordance with the AXE concept or the company exchange MD-110, and other companies that market similar switching centres and exchanges are mainly utilized through the medium of the telephone keypad. The keypad is fixedly mounted to the telephone base unit to which the receiver of the telephone apparatus is connected by a lead. It is therefore necessary for a person wishing to use the telephone to go to the base unit and initiate a desired service through the keypad.

Cordless telephones with which one can move freely in relation to a base unit during a call are common apparatus today. However, the cordless telephones per se include all of the functions possessed by a base unit equipped with a conventional telephone receiver, e.g. keypad, dialling function, radio part, etc. , wherein the base unit includes a radio receiver and units for converting radio conversation signals to telephone signals for telephony on line networks.

Also known to the art are acoustic devices for setting up telephone calls, wherein a DTMF transmitter (Dual Tone Multi Frequency transmitter (dialling)), for instance, transmits acoustic signals by virtue of holding the device close to the microphone in the telephone receiver when initiating a call setup.

With regard to telephone conversations made on a loudspeaker telephone, a telephone receiver is unnecessary except as an alternative. The need to go to a base unit and initiate a telephony service through the medium of the keypad and to converse with the aid of a telephone receiver, as in present- day cases, is therefore essentially superfluous, and there is a desire to be able to initiate a call setup and a call release from anywhere in a room provided with a loudspeaker telephone. Devices provided with remote keypads which solely transmit signals for setting up and releasing calls through the medium of infrared technique or radio technique are as yet still unknown in the present technical field.

SUMMARY OF THE INVENTION

One object of the present invention is to provide a sound unit which is not troubled by acoustic feedback. Another object of the invention is to provide for adaptive extinguishing of echoes.

A further object of the invention is to provide sound units that can be mutually connected and placed in one and the same room, wherein a sound unit constitutes a central unit in which no acoustic feedback is generated.

Still another object of the invention is to provide sound units that can be placed essentially anywhere in a room, for instance on any supportive surface such as a conference table, a wall, against a ceiling, etc. This can be achieved in all of said places in a room when several inventive sound units are connected together and where one sound unit is a central unit. Sound units may also be arranged in separate rooms.

Yet another object of the invention is to provide a control function which controls adaptive extinguishing of acoustic feedback and echoes in an arrangement having at least one sound unit.

The aforesaid enables conference members to move freely in the room during an ongoing conference and to speak from any chosen direction in relation to the positioning of the microphones and with immediate speech exchange between conference members while achieving the aforesaid objects.

A further object of the present invention is to provide a device which includes a remote keypad and by means of which call connections can be set up and released. Such a device solves the problem of being unable to move freely in a room equipped with a loudspeaker telephone in order to initiate a call setup and call release.

In a preferred embodiment , the invention is practiced with a voice communication device for recording and transmitting voices and replaying said voices in full duplex. In this regard, the device includes voice communication means and adaptive signal processing means having an interface with the voice communication means for cancelling echo and acoustic feedback. Voices are recorded with the aid of at least one all-directional microphone which is placed centrally in relation to at least two loudspeakers. The at least two loudspeakers are positioned to surround the microphone, essentially in the immediate vicinity thereof, and together with the microphone form a voice-recording and voice-replay¬ ing sound unit which is connected to the adaptive signal processing means.

The voice communication means includes equipment for estab- lishing and maintaining telephone call connections, wherein the device is intended for transmitting telephone calls, either with or without a telephone receiver, via the voice communication means including loudspeaker functions for conference calls, wherein said interface is a line interface against the equipment functioning to establish and maintain telephone call connections.

In a further embodiment of the invention, the voice communi¬ cation device is intended for voice-recording purposes, wherein the voice communication device includes sound recording equipment for playing-in recorded voices, for instance in recording sound during film recordings, said interface being an interface for adaptation to the sound recording equipment.

It is possible to connect together several of the aforesaid sound units, wherein one of the sound units will include the voice communication means and the adaptive signal processing means, this unit forming a so-called central sound unit. Mutually interconnected sound units are placed appropriately in one or more rooms, preferably on a supportive surface such as a wall or on the ceiling of a room in mutually spaced relationship.

In a preferred embodiment, each microphone is surrounded by three loudspeakers, wherein the loudspeakers for each individual sound unit are placed symmetrically around the microphone of the sound unit.

The adaptive signal processing means comprises digital signal processing means (DSP) with controlled signal processing. The signals are processed by a digital signal processor (DSP) which includes three main components designated room echo cancellers, line echo cancellers and echo cancelling control means. The cancelling means are comprised of adaptive transversal filters having time lengths which correspond to the principle echo normally experienced in office environ¬ ments or in conference rooms and the electric echo experi¬ enced on an analog telephone line respectively, said filters adapting to occurrent echoes through the medium of error signals in accordance with the least squares method.

The room echo canceller and the line echo canceller are connected on the input side to outgoing lines and incoming lines respectively and deliver on their respective outputs output signals which are subtracted in subtraction means from incoming room noise and incoming line noise respectively. Incoming room noise and incoming line noise is attenuated by voltage-controlled amplifiers for stabilizing incoming sound, this sound being delivered from the amplifier outputs to the line echo canceller and the room echo canceller respectively and also to the interface and loudspeaker unit respectively.

The echo cancelling control means effects a backup control of the voltage-controlled amplifiers in time with respective speech signals, and also a backup control of the room echo canceller and the line echo canceller.

The voice communication device also transmits so-called pink noise when establishing a call connection.

The present invention also relates to a method of recording and transmitting voices and playing-back said voices in full duplex. The method includes the use of voice communication means and adaptive signal processing means (DSP), with interfaces against the voice communication means, and echo and feedback cancelling means. Voices are recorded by at least one microphone which is placed centrally in relation to at least two loudspeaker units and which records from all sides, i.e. is all-directional. The microphone is surrounded by at least two loudspeaker units placed essentially in the immediate vicinity of the microphone, and the loudspeakers and microphone form a sound unit for recording and playing- back voices and connected to the adaptive signal processing means.

The signals are processed by a digital signal processor (DSP) which includes three principle means designated room echo canceller, line echo canceller and echo cancelling control means. The cancelling means are comprised of adaptive transversal filters having lengths in time that correspond to the principle echo occurring in typical office environ¬ ments or in a conference room, and the electric echo that occurs on an analog telephone line. The filters adapt to occurrent echoes through the medium of error signals in accordance with the least squares method. The room echo canceller and the line echo canceller are connected on their input sides to respective outgoing lines and incoming lines.

The outputs of the cancellers deliver output signals which are subtracted in subtraction means from incoming room sound and incoming line sound respectively. Incoming room sound and incoming line sound is attenuated by voltage-controlled amplifiers for stabilizing incoming sound, wherein the sound is delivered from the amplifier outputs to the line echo canceller and the room echo canceller respectively, and also to the interface and loudspeaker unit respectively. The means for controlling the function of echo cancellation forms a backup control for the voltage-controlled amplifiers in time with respective speech signals, and also a backup control for the room echo canceller and the line echo canceller.

The present invention also relates to a device having a remote keypad for loudspeaker telephones having keys or buttons which exhibit those functions that are required to implement a telephone exchange service and a telephone function service in conjunction with setting up and releasing a call. The keypad includes coding means which are connected to the buttons by conductor paths. Depression of the buttons, or keys, is converted by signal generating means into pulse trains for cordless transmission of coded signals to receiver means for receiving generated signals in a receiver unit. The receiver unit is arranged in a central unit for loudspeaker telephony.

The receiver unit is controlled and monitored by a processor unit which decodes and converts generated signals to tele¬ phone signals adapted to the connected telephone network. With the aid of control signals, the processor coordinates the functions initiated by a user via key depressions at a distance from the central unit.

Generated signals are electromagnetic or acoustic.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be described in more detail with reference to preferred embodiments thereof and also with reference to the accompanying drawings, in which

Fig. 1 illustrates schematically a sound unit for a confer¬ ence telephone in accordance with known techniques; Fig. 2 illustrates a voice-recording unit schematically from above;

Fig. 3 is an exploded view of a voice-recording unit;

Fig. 4 is a block schematic illustrating the adaptive digital signalling process in accordance with a particular embodiment of the invention;

Fig. 5 illustrates diagra matically the arrangement of three loudspeaker units and shows the sound area and the arrange¬ ment of an all-directional microphone;

Fig. 6 illustrates a conference telephone system including remote manoeuvred keypads;

Fig. 7 illustrates a remote keypad and shows examples of key- related functions; and

Fig. 8 illustrates a remote manoeuvred key sender with a receiver unit in accordance with one embodiment of the invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS OF THE INVENTION

The present invention and its preferred embodiments described below is/are based on a novel and innovative positioning of loudspeakers and microphones. It has not earlier been possible to place a microphone centrally with several loudspeaker units positioned around the microphone and in its immediate vicinity. Such an arrangement leads to the occur¬ rence of loudspeaker acoustic feedback via the microphone, so-called acoustic feedback instability or singing. In accordance with the present invention, this problem is solved by adaptive (automatic) continuous control of occurrent acoustic feedback, line echoes and changes in the acoustics of a room through, the medium of modern digital signal processing technology, so-called "Digital Signal Processing" (DSP) .

The invention lays no claim to telecommunications telephony techniques known to the person skilled in this art, and such techniques will not therefore be described in detail. On the other hand, the invention does lay claim to voice-recording devices that refer to the arrangement of microphones and loudspeakers for achieving the best possible sound recording and playback quality, and to the signal processing carried out in conjunction therewith. The present invention particu¬ larly relates to a voice-recording device with known tele¬ phone techniques for loudspeaker telephones, preferably for conference calls. However, the invention does not exclude the use of the voice-recording device in sound recordings, for instance in recording sound tracks on film, music, etc., to obtain a uniform sound quality over a wide area.

Fig. 1 illustrates very schematically a sound unit 10 having a casing 12 and intended for recording voices in telephone conferences with an arrangement of microphones 14 and loudspeaker units 16 in accordance with a technique which is unique as far as we are aware. Acoustic feedback is avoided by placing three directionally appointed microphones 14 around a centrally placed loudspeaker unit 16 at an angular spacing of 120^*. The microphones are preferably directionally appointed so that the sound recorded thereby will not overlap and thus generate acoustic feedback and undesirable echoes, for instance by sound propagation and rebound such as to have a pronounced time delay when reaching some of the micro^¬ phones.

An arrangement according to Fig. 1 must be supported, for instance on a table, wherein the conference members will preferably be uniformly placed around the arrangement. The aforesaid problems render the known arrangement unusable if it is suspended from a wall or from a ceiling.

Fig. 2 illustrates an inventive voice-recording device 20 schematically from above. The device includes an outer casing 22 having three groups of gratings or grids 24 for the passage of sound from three loudspeaker units 16, slots 26 for the passage of sound to an all-directional recording microphone 28, see Fig. 3, lamps in the form of light- emitting diodes 30 for indicating when the device is switched-on, and a keypad area 32.

An exploded view of a voice-recording device 20 is given in Fig. 3, from which the construction of the device can be seen in more detail. In addition to what is shown in Fig. 2, the Fig. 3 illustration illustrates the all-directional recording microphone 28, the keypad 34 having two function keys 36 - and + for controlling the sound from the loudspeaker units 16. Also shown is an on/off button or key 38, and a privacy button or key 40 on the keypad 34.

Also illustrated schematically is the circuit board construc¬ tion 42 including means having components and using known techniques for establishing and maintaining telephone call connections, with built-in loudspeaker functions. The circuit board construction 42 also includes means having components for the innovative adaptive control function for cancelling echoes and acoustic feedback. Also shown is the bottom part 44 of the device 20 connected to the outer casing 22.

Fig. 4 illustrates function blocks and signal diagrams relating to echo-cancelling means. In the illustrated case, the signal processing means with echo cancellation is comprised of a powerful digital signal processor (DSP), framed in broken lines in the Figure. The DSP unit as disposed in accordance with the present invention is com¬ prised of three principle means shown as blocks in the Figure, wherein the block REK indicates a room echo cancel¬ ler, the block LEK indicates a line echo canceller and the block RFOE indicates echo-cancelling control function means. The signal output from the REK means is connected to a subtraction circuit 46. Similarly, the signal output from the LEK means is connected to a further subtraction circuit 48.

With the intention of providing a better understanding of the signal flow into and out of the DSP unit, a definition is given in the following of the acoustic and electric signal designations used, wherein Atal designates the speech or acoustic signal of the A-subscriber (i.e. the person speaking in the room in which the voice-recording device 20 is located, in this case the conference telephone and optionally additional connected sound units), Btal designates corre¬ spondingly the speech of a B-subscriber arriving from another telephone at the other end of a call connection, RAtal designates reflected Atal, RBtal correspondingly designates reflected Btal, rAtal designates estimates reflected Atal in accordance with a mathematical model included in the REK means, rBtal designates correspondingly estimated reflected Btal included in the LEK means. It is appropriate in the following to consider Atal and Btal as alternating acoustic and digital electric signals respectively, since the inven- tion pertains to the conversion of sound to digital signals (optionally with later conversion to analog electric signals) and vice versa.

The DSP unit includes a further two units designated Gu and Gi respectively, said units constituting speech amplify^¬ ing/speech attenuating means. The embodiment according to Fig. 4 is shown to exemplify adaptive echo cancelling in accordance with the invention. For the sake of illustration, only one loudspeaker unit 16 is shown. Also shown in the Figure is an object 50, for instance a wall, from which sound is reflected. Btal reflected by the object 50 is indicated by two arrows. The cancelling means REK and LEK in the DSP unit are com¬ prised of adaptive transversal filters having time lengths which correspond to the principle echo in a typical office environment or a typical conference room, and the electric echo on an analog telephone line respectively. The filters are updated, i.e. adapted to occurrent echoes, via the error signals ε_r and t_x according to the classic Least Mean Square algorithm (the LMS algorithm) where index r denotes space and index 1 denotes lines, to refer to the direction of signal propagation in the DSP unit. In order for the LMS algorithm to work optimally, i.e. to follow changes in room acoustics as quickly as possible, without being disturbed by simulta¬ neous Atal and Btal, the LMS algorithm has been developed with backup control from the RFOE means, the control signals to the means REK and LEK being marked with broken arrows in the Figure. Backup control, which also controls the Gu and Gi units whose control signals are marked with broken arrows in the Figure, has decisive significance in the speech quality and speech stability experienced. The Gu and Gi units are voltage-controlled amplifiers which dampen incoming signals with 6 dB, which is not audible to the human ear and which will thus have no appreciable influence on outgoing and incoming Atal and Btal signals respectively. Attenuation levels that are audible to the human ear lie approximately in the range of 12-20 dB.

Backup control is effected in accordance with the following criteria:

- Error signals in the REK means for sound emanating from the room are only updated in respect of Btal or strong Btal + weak Atal, i.e. when Atal and Btal respectively are detected. The same applies to the error signals in the LEK means.

Seek optimal convergence factors as a function of ε_r and t_x to obtain rapid adaptation to changes. The convergence factor shows how much of ε_r and ε_. is used for updating REK and LEK respectively. The factor is a known factor for LMS algorithms with which error signals are multiplied and determines the speed at which said model in the DSP unit is adapted.

Avoid adapting to ε_r and ε__. which include pure tones (i.e. without harmonics) since this can lead to divergence.

- Prevent the over-excitation or over-modulation of loud¬ speaker units, which would otherwise result in non-linear and non-cancellable echoes (RBtal, RAtal).

Lower the convergence factor at high input levels with risk of moderate distortion.

The so-called all-directional recording microphone 28 picks up Atal + RBtal and converts these acoustic signals to electric signals and delivers said signals on the line 52, whereafter rBtal is subtracted from the added signals in the subtraction circuit 46, said rBtal being estimated in the REK means in accordance with the aforegoing. The subtracted signal, now comprised of Atal + ε_r is attenuated or amplified in the Gu unit in time with Atal so as to further enhance the stability of the signal. Atal + ε_r is delivered from the Gu unit to the LEK means, which estimates rAtal in accordance with the aforegoing. The signal is also delivered to a line interface 47 for adaptation to telephony equipment in accordance with known techniques, this line interface, in turn, delivering Atal to the telephony equipment on a two- wire line 56.

Similarly the line interface receives Btal mixed with Atal. The signal applied to the line 54 is fed into the DSP unit from the line interface 47, whereafter rAtal, which has been estimated in the REK means according to the aforegoing, is subtracted from the signal in the subtraction circuit 48. The subtraction signal is now comprised of Atal + _x and is attenuated in the Gi unit in time with Btal so as to further enhance signal stability. Btal + ε_r is now delivered from the Gi unit to the REK means, which estimates rBtal in accordance with the aforegoing. The signal is thereafter delivered via the loudspeaker unit 16 as audible Btal, whereafter the procedure is repeated cyclically during ongoing conversations with adaptive echo cancelling.

As be before mentioned, the adaptive signal processing means is comprised of a powerful DSP, which is able to compile in its memory models for both acoustic echoes in space (REK) and electric line-echoes (LEK) . Atal and Btal are mutually separated by these echoes in the DSP means, so as to prevent acoustic feedback at normal speech levels (about 70-75 dB) . A stable function and optimal setting of the cancellers REK and LEK is an absolute requisite for good duplex-speech quality in the conference telephones. This is accentuated by virtue of the strong acoustic coupling according to the geometry in Fig. 4, i.e. the risk of so-called acoustic feedback is great. Cancelling, or extinguishing, of the echoes is achieved by:

Subtracting from the sound incoming to the microphone - Atal + RBtal rBtal = RBtal + ε_r, such as to result in Atal + ε_r, i.e. pure Atal apart from a room residual = e_r.

Correspondingly cancelling on the line side, therewith also reducing the risk of acoustic feedback and preventing part of Atal being reflected back into the room via loudspeaker units 16.

The inventive device also sends out pink noise on the line during a call connection, to extinguish noise and to quickly stabilize call quality prior to the adaptive control accord¬ ing to the invention taking over. The innovative arrangement is illustrated in Fig. 5, wherein three loudspeaker units 16 surround an all-directional microphone 28, i.e. a microphone which picks up sound from a 360^* area, wherein each of the loudspeaker units 16 covers a sound-recording area of 120^*.

Remote control of ringing functions on a modern loudspeaker conference telephone will now be described with reference to Figs. 6-8 in accordance with the present invention, wherein Fig. 8 includes the designations given in Figs. 2-5 with related text.

Fig. 6 illustrates a loudspeaker telephone system having a central unit 60 which is connected to two ceiling-mounted sound units 62. The illustrated device covers a large conference room with regard to voice-recording. In this regard, only the central unit 60 has all functions necessary for dialling and maintaining a call connection, wherein the sound unit 62 includes the unit of Fig. 5 with loudspeaker units 16 and an all-directional recording microphone 28, and necessary means for transmitting recorded sound to the central unit 60 and from there to the telephone network.

The embodiments of the invention will be exemplified herein- after with reference to the use of infrared radiation for cordless transmission of control functions. However, this does not exclude the use of radio wave technology, ultrasound technology or laser technology for cordless transmission of control functions in practical applications.

The central unit 60 is supplied with operating voltage from the net transformer unit 69, for instance of the kind marketed by Tufvasson and designated PFLF15. The designation 68 relates to the connection to the telephone network via a jack and a jack wall connection. An infrared receiver (IR receiver) 66 located on the central unit 60 receives IR- radiation emitted from the keypad 70. The central unit 60 also includes a simplified keypad or button pad 34 having an on/off function, sound level control keys and a green light- emitting diode for indicating the on/off state.

The keypad 70 illustrated in Fig. 7 has the functions normally found on a conventional keypad telephone which has a loudspeaker function. The keypad shown in Fig. 7 switches the central unit 60 on and off and also possible other sound units in accordance with Fig. 6, via the button or key 71. When the central unit 60 is switched on, it generates noise of short duration in the loudspeakers. This is the intrinsic tuning of the system. With the aid of the noise in the loudspeaker units, the system including the central unit 60 and possibly other sound units, senses the acoustic in a conference room and adapts itself accordingly. The privacy key 72 disconnects/connects the microphones and is also used to exclude the dialling of a number which includes the digit/numerical keys 76, 0-9. The keys 73 are used to adjust the volume of a conversation. The keys 73 also include the additional function of enabling the central unit 60 to be manually trimmed by pressing both keys simultaneously, whereupon noise of short duration is heard in accordance with the aforegoing. In the case of satellite connected calls, manual trimming will preferably be effected subsequent to the called party having answered the call. Alternatively, the central unit can be trimmed manually with the aid of an additional key or button intended for this function (not shown). Additional keys or buttons will preferably be provided for storing card numbers (not shown). The keys need not necessarily be press buttons, but can be any type of key sets that are conceivable or commercially available, such as touch pads, for instance.

The reference numeral 74 designates the asterisk, square and R keys. All services provided in modern telephone exchanges can be used through the medium of these keys. Examples of such telephone exchanges are AXE stations and company exchanges, e.g. MD-110 from Ericsson. Repetition of the number last entered is achieved with the [=] key 75.

Fig. 8 illustrates the cordless transmission of digital IR pulse trains 80 from the keypad 70 to the receiver unit 82 located in the central unit 60. The keypad 70 includes a key matrix 84 which forms a closed current path with the broken- line conductor parts 86 and the broken-line horizontal conductor parts 88 when depressing a key or button 84. The thus closed current path addresses an encoder also included in the keypad 70, wherein the encoder delivers a coded IR pulse train 80 corresponding to the depressed key 84 via the transmitting IR diode 89. It is assumed in this case that the encoder includes circuits for conversion to IR signals, such as integrated circuits marketed by different manufacturers.

The IR pulse train 80 is received cordlessly by an IR radiation receiving IR diode 90 and is transmitted to an IR receiver in which the IR pulse train 80 is converted to digital signals which are decoded in a connected decoder. The receiver unit 82 includes a PIC processor (Peripheral Interface Controller processor, single chip computer) which controls and monitors the circuits in the receiver unit 82). The processor processes digital signals incoming from the decoder and relating to the numerical keys 0-9, * (asterisk) and # (square) keys and transmits these signals to a DSP unit (Digital Signal Processing unit) in which the decoded signals are converted to DTMF signals (Dual Tone Multi Frequency dialling signals) for transmission to a telephone exchange via the telephone network for the purpose of setting up and releasing a call connection. The DSP unit of the Fig. 8 embodiment should not be confused with the DSP unit of the Fig. 4 embodiment, this latter DSP unit being a free-standing unit for cancelling line echoes and acoustic feedback. The processor also controls control logic which deliver control signals necessary for synchronizing mutually coacting functions when setting up and releasing a call connection. The signal controlling the R-key function (Register recall) is delivered via control logic which produces an interruption of about 100 msek. The R-key function does not deliver a DTMF signal, but initiates an interruption via the control logic. Other keys that use the control logic are, for instance, volume, microphone on/off (mute) and the trimming function.

It is known to remotely control television apparatus with IR radiation, among other things. This function is also known in connection with cordless telephones, although these telephones also transmit speech signals to a base unit, wherein all telephony functions are integrated in the cordless transmitting telephone. The base unit thus converts only the transmitted signals, preferably radio signals, to adapt said signals to the telephone network. However, telephony applications are not earlier known in which solely the keypad is a free-standing unit so as to enable connec¬ tions to be set up and released at distances from the actual central unit of the telephone. There are innumerable examples of where such a feature is desirable, of which one such example is when a speaker gives a lecture and some person located in some other place needs to be consulted so that all lecture participants can hear the conversation. The lecturer need not then interrupt the lecture, but is able to set up a call with a loudspeaker function quickly and simply from his/her lecturing position.

Although the present invention has been described with reference to preferred exemplifying embodiments thereof, it will be understood that these embodiments do not limit the scope of the invention but merely illustrate the invention to one of normal skill in this art. The invention is thus restricted solely by the scope of the following Claims.

Claims

1. A voice communication device (20) for recording and transmitting voices and replaying said voices in full duplex, characterized in that the device (20) includes voice communi¬ cation means and adaptive signal processing means (DSP), with an interface (47) against the voice communication means, for cancelling echo and acoustic feedback, wherein said voices are recorded by at least one so-called all-directional recording microphone (28) arranged centrally in relation to at least two loudspeaker units (16), and wherein said at least two loudspeaker units (16) are arranged to surround the microphone (28) essentially in the immediate vicinity of the microphone and together with the microphone form a voice- recording and voice-replaying sound unit which is connected to said adaptive signal processing means.

2. A device according to Claim 1, characterized in that the voice communication means includes telephone call establish- ing and maintaining equipment.

3. A device according to Claim 2, characterized in that the interface (47) is a line interface against said telephone call establishing and maintaining equipment.

4. A device according to Claim 3, characterized in that said device is intended for the transmission of telephone calls, with or without a telephone receiver, through the medium of the voice communication means that includes a loudspeaker function.

5. A device according to Claim 4, characterized in that said device is intended for conference calls.

6. A device according to Claim 1, characterized in that said device is intended for voice recording, wherein the voice communication means includes sound-recording equipment for playing-in recorded voices.

7. A device according to Claim 6, characterized in that the interface is an interface for adaptation to said sound- recording equipment.

8. A device according to any one of the preceding Claims, characterized in that several sound units can be connected to one another, wherein one of said sound units includes said voice communication means and said adaptive signal processing means.

9. A device according to Claim 8, characterized in that sound units that are connected to one another are disposed in suitable places in one or several rooms, preferably on a supportive surface, such as a wall or a ceiling, in mutually spaced relationship.

10. A device according to any one of the preceding Claims, characterized in that each microphone is surrounded by three loudspeakers.

11. A device according to any one of the preceding Claims, characterized in that the loudspeakers of each individual sound units are placed symmetrically around the microphone of said sound unit.

12. A device according to any one of the preceding Claims, characterized in that the adaptive signal processing means is a digital signal processing means (DSP) .

13. A device according to Claim 12, characterized in that signal processing is carried out by a digital signal proces¬ sor (DSP) that includes three principle means designated room echo canceller (REK), line echo canceller (LEK) and echo cancelling control function means (RFOE) .

14. A device according to Claim 13, characterized in that the cancelling means (REK, LEK) are comprised of adaptive transversal filters having time lengths that correspond essentially to the echo that occurs in typical office environments or in a typical conference room, and the electric echo that occurs on an analog telephone line respectively, wherein the filters adapt themselves to occurrent echoes through the medium of error signals (ε_r, ε _x ) in accordance with the least square method.

15. A device according to Claim 14, characterized in that the room echo canceller (REK) and the line echo canceller (LEK) are connected on the input side to a respective outgoing line (52) and incoming line (54), wherein the cancellers deliver output signals (rBtal, rAtal) on their respective outputs, said output signals being subtracted from incoming room sound (Atal + Btal) and incoming line sound (Btal + Rtal) respectively in respective subtraction means (46, 48).

16. A device according to Claim 15, characterized in that incoming room sound and incoming line sound is attenuated by voltage-controlled amplifiers (Gu, Gi) to stabilize incoming sound, said sound being delivered from the amplifier outputs to the line echo canceller (LEK) and the room echo canceller (REK) respectively, and to the interface (47) and the loudspeaker unit (16) respectively.

17. A device according to Claim 13, characterized in that the echo cancelling control function means (RFOE) forms a backup control for the voltage-controlled amplifiers (Gu, Gi) in time with respective speech signals (Atal, Btal) and a backup control for the room echo canceller (REK) and the line echo canceller (LEK), wherein said backup control is effected by: updating error signals in the REK means for room sound solely when Btal or strong Btal + weak Atal occurs, i.e. when Atal and Btal respectively are detected, wherein corresponding updating is effected in respect of the error signals in the LEK means; seeking optimal convergence factors as a function of the error signals (ε_r, t_ ) to obtain rapid adaptation to speech changes, wherein the convergence factor discloses how large a part of the error signals (ε_r, ε_.) were used in updating REK and LEK respectively; avoiding the adaptation of error signals (ε_r, ε __) that contain pure tones; preventing overmodulation or overexcitation of loud¬ speaker units; and lowering the convergence factor at high sound levels with risk of moderate distortion.

18. A device according to any one of Claims 1-5 and Claims 8-17, characterized in that the device is adapted to transmit so-called pink noise setting up a call connection.

19. A method for recording and transmitting voices and replaying said voices in full duplex, characterized in that the method comprises the use of voice communication means and adaptive signal processing means (DSP) having an interface (47) against said voice communication means, for cancelling echo and acoustic feedback, wherein voices are recorded by means of at least one all-directional recording microphone (28) arranged centrally in relation to at least two loud¬ speaker units (16), wherein said at least two loudspeaker units (16) are disposed so as to surround the microphone (28) essentially in the immediate vicinity thereof and, together with said microphone, form a voice-recording and voice- replaying sound unit which is connected to said adaptive signal processing means, wherein signal processing is carried out by a digital signal processor (DSP) which includes three principle means designated room echo canceller (REK), line echo canceller (LEK) and echo cancelling control function means (RFOE), wherein the cancelling means (REK, LEK) are comprised of adaptive transversal filters having time lengths which correspond to the principle echo occurring in typical office environments or in typical conference room environ¬ ments, and the electric echo occurring on an analog telephone line respectively, wherein the filters adapt themselves to occurrent echoes through the medium of error signals (ε_r, ε in accordance with the least squares method, wherein the room echo canceller (REK) and the line echo canceller (LEK) are connected on the input side to a respective outgoing line (52) and an incoming line (54), wherein the cancellers deliver on their respective outputs output signals (rBtal, rAtal) which are subtracted from incoming room sound (Atal + RBtal) and from incoming line sound (Btal + Rtal) in subtraction devices (46, 48), wherein incoming room sound and incoming line sound are attenuated by voltage-controlled amplifiers (Gu, Gi) for stabilizing incoming sound, said sound being delivered on the amplifier outputs to a respec¬ tive line echo canceller (LEK) and room echo canceller (REK) and respectively to the interface (47) and the loudspeaker unit (16), wherein the echo cancelling control function means (RFOE) carries out a backup control for the voltage-con¬ trolled amplifiers (Gu, Gi) in time with respective speech signals (Atal, Btal), and a backup control for the room echo canceller (REK) and the line echo canceller (LEK), and wherein the backup control includes the steps of: updating error signals in the REK means for room sound solely when Btal or strong Btal + weak Atal occurs, i.e. when Atal and Btal respectively are detected, wherein corresponding updating is effected in respect of the error signals in the LEK means; seeking optimal convergence factors as a function of the error signals (ε_r, t _) to obtain rapid adaptation to speech changes, wherein the convergence factor discloses how large a part of the error signals (ε_r, e_ ) was used in updating REK and LEK respectively; avoiding the adaptation of error signals (ε_r, t _) that contain pure tones; preventing over odulation or overexcitation of loud¬ speaker units; and lowering the convergence factor at high sound levels with risk of moderate distortion.

20. A device including a remote keypad (70) for loudspeaker telephones having keys which exhibit those functions required to utilize a service and function offered by a telephone exchange in conjunction with setting up and releasing a call connection, characterized in that the keypad includes coding means connected to the keys (84) of said keypad by connec¬ tions (86, 88), wherein key depressions are converted via a signal generating means (89) to a pulse train (80) for cordless transmission of coded signals to receiver means (19) for receiving generated signals in a receiver unit (82) arranged in a central unit (60) for loudspeaker telephony, said receiver unit (82) being controlled and monitored by a processor (PIC processor) which, via a connected decoder, decodes and converts generated signals to telephone signals through the medium of signal processing means (DSP) adapted to the connected telephone network, and which, through the medium of control signals from control logic, controls and coordinates those functions that are initiated by a user at a distance from the central unit (60) by depressing said keys.

21. A device according to Claim 20, characterized in that generated signals are electromagnetic.

22. A device according to Claim 20, characterized in that generated signals are acoustic.

23. A device according to Claim 20, 21 or 22, characterized in that the central unit (20) includes voice communication means and adaptive signal processing means, including an interface (47) against the voice communication means, for cancelling echoes and acoustic feedback, wherein voices are recorded by means of at least one all-directional recording microphone (28) arranged centrally in relation to at least two loudspeaker units (16), wherein said at least two loudspeaker units (16) are disposed so as to surround the microphone (28) essentially in the immediate vicinity thereof, and wherein the loudspeaker units and the microphone together form a voice-recording and voice-replaying sound unit which is connected to the adaptive signal processing means.

24. A device according to Claim 23, characterized in that the voice communication means includes telephone call establishing and maintaining equipment.

25. A device according to Claim 24, characterized in that the interface (47) is a line interface against said telephone call establishing and maintaining equipment.

26. A conference telephone device (20) for recording and transmitting voices and replaying said voices in full duplex, characterized in that the device (20) includes voice communi¬ cation means and adaptive signal processing means (DSP), and an interface (47) against said voice communication means, for cancelling echoes and acoustic feedbacks, wherein voices are recorded by means of at least one all-directional microphone (28) disposed centrally in relation to at least two loud¬ speaker units (16), wherein the at least two loudspeaker units (16) are disposed so as to surround the microphone (28) essentially in the immediate vicinity thereof, and wherein the loudspeaker units and the microphone together form a sound unit for conference telephony and for voice playback connected to the adaptive signal processing means.