WO1990001843A2

WO1990001843A2 - Remote dictation system using telephone line

Info

Publication number: WO1990001843A2
Application number: PCT/GB1989/000864
Authority: WO
Inventors: John Alfred Edwards
Original assignee: John Edwards Technology Group Limited
Priority date: 1988-07-29
Filing date: 1989-07-28
Publication date: 1990-02-22
Also published as: PT91313A; GB8818170D0; AU4041189A; WO1990001843A3

Abstract

The communications system enables a user to dictate a message at a subscriber station (2) and the message is typed by an operator at an operator station (6) so that the text of the message can be displayed at the subscriber station. A notch filter (3, 7) enables voice and text data to be communicated simultaneously over the same telephone channel.

Description

REMOTE DICTATION SYSTEM

USING TELEPHONE LINE

This invention relates to a communications system. The system enables a user to dictate a message at a subscriber station which is typed by an operator at an operator station so that the text of the message can be displayed at the subscriber station.

A well known disadvantage of computers is that they require keyboard skills to enable operation by a user. Although the lack of skill does not prevent a user from making simple entries on a keyboard to control the computer, the unskilled user would find it difficult and time- consuming to use the keyboard to enter lengthy passages of text, for example, in the case of wordprocessing, facsimile messaging and similar applications. In such circumstances, it is usually necessary to employ a skilled typist or keyboard operator so that work can be carried out at a reasonable speed and with sufficient accuracy. However, the demand on a skilled persona time may be periodic rather than continuous and sometimes required outside the normal working hours. There is therefore a need for a service which would enable text to be created from a dictated message as and when there was a need for a skilled operator. The invention seeks to solve this problem.

In accordance with the invention, a method of communication between a subscriber station and an operator station is characterised by the steps of: a) dictating a voice message from the subscriber station to the operator station over a voice channel, b) generating keyboard signals at the operator station corresponding with characters in text representing the voice message, and c) causing respective computers at each station to operate so that data representing the keyboard signals is transmitted from the operator station to the subscriber station over a data channel to enable the keyboard signals to be displayed at the subscriber station. Also in accordance with the invention, a communications system comprises: a subscriber station and an operator station; a respective computer at each station, each computer including a display for viewing text created by digitised keyboard signals stored in a respective buffer, a keyboard being provided at least at the operator station; a data channel enabling data to be exchanged between the computers; a voice channel enabling voice communication between the stations; either said data arid said voice channels being independent, or means being provided to enable the data to be exchanged and the voice communication to occur during the same call; characterised in that said computers are programmed to cause the system to operate so that each keyboard signal is encoded at the operator station in a manner in which it can be decoded and recognized as a keyboard signal at the subscriber station whereby, as the subscriber dictates a voice message over the voice channel, data representing each character keyed by the operator at the operator station is transmitted over the data channel and decoded so that it can be viewed on the display at the subscriber station. Besides the advantage of enabling an unskilled user to have text displayed in response to a dictated message at any convenient time, the invention also has the advantage that no serious loss of text will occur when, for example, there is a break in transmission. This is due to the fact that each character is typed, transmitted and stored for viewing as it is generated. This differs considerably from known

•electronic mail¹ wherein blocks of text are stored at an originating data terminal, transmitted (en bloc) to a mainframe computer and subsequently distributed to each receiving data terminal or electronic mailbox.

Preferably, the arrangement is such that when each key is struck on the operator's keyboard, this causes an appropriate scan code and ASCII code to be generated in the usual way to produce a digitised signal. However, the interrupt due to the keystroke is used to trigger a sub¬ routine to split the encoded and digitised signal into parts which are further encoded with. Hamming codes to ensure error-free decoding after transmission. Data transmission over public networks can be corrupted by noise and this can lead to errors in the display of text at the subscriber station. The resultant data segment, which represents a keyed character, is also prefixed with a recognition code to which the computer at the subscriber station responds as if it were a locally generated keystroke interrupt. The computer at the subscriber station is programmed to decode the received signal so as to reproduce the digitised signal which is stored in the buffer as a standard character signal for display. Coded portions of the data segment and the prefix are preferably doubled, prior to transmission, to ensure receipt of at least one portion that is not corrupted by noise.

Preferably, both voice and the data channels share the same communications link. This requires the use of special modems respectively provided for transmitting and receiving digitised characters over the communications link. In a preferred form of modem, filters are provided so as to create a notch in the voice band of the voice channel, the notch enabling the data channel to operate simultaneously within the voice channel. This enables the invention to be used with a public switched telephone network or a radio link.

Alternatively, the voice and data channels are independent of one another. For example, a telephone connection is made via a PABX, intercom or direct line. The computers may be connected by a data channel and a line driver and a selector switch can be used, if required, in the data channel. Since the channels are independent, they can be used simultaneously for the purposes of the invention.

Other alternatives are to use either a digital telephone connected to a fifth generation 2B+D PABX having an RS 232C serial port for entering a data stream, or an Integrated Switched Data Network (ISDN) having an integral RS 232C serial data interface port.

Embodiments of the invention will now be described with reference to the accompanying drawings in which:

Fig. 1 illustrates a first -embodiment of the invention which employs, for example, a public telephone network, modems and filters to enable simultaneous voice and data communication, Fig. 2 illustrates the characteristics of the filters system used in Fig. 1,

Fig. 3 illustrates a second embodiment of the invention employing independent data and voice channel, and

Fig. 4 illustrates a third embodiment which employs a 2B+D PABX.

Referring to Figs. 1 and 2, the first embodiment of the invention comprises a subscriber telephone 1 and a computer system 2 connected, via a unit 3, to a Public Switched Telephone Network (PSTN) 4, or radio link (not shown). An operator telephone 5 and computer system 6 are similarly connected to a unit 7 and to the network or link 4.

Each computer system 2, 6 includes a computer 2a, 6a, a keyboard 2b, 6b and a display screen 2c, 6c. A printer 2d is optionally connected to computer 2b. The operator's telephone 5 includes a headset 5a.

Each of the units 3,^' 7 comprise a notch filter 3a, 7a connected to the respective telephone 1, 5; a data modem 3b, 7b for connection to the respective computer system 2, 6; a filter 3c, 7c having a pass band which fits substantially within the notch of the telephone filter 3a, 7a; and a line interface 3d, 7d for interfacing the outputs of the filters with the PSTN 4. The data modems 3c, 7c are connected to their respective computers 2a, 6a by means of RS 232C serial ports.

Fig. 2 illustrates the characteristics of the filters in each unit 3, 7. The telephone filter has a notch in the speech band between approximately 1700 hz and 2700 hz into which the data pass band is received with cross-overs at approximately -40 dBm. The graph also shows a flat plateau shape to the line characteristics over most of the 0-4000 hz frequency band. The advantage of using a data channel within a notch in the voice band is that the data transmission is far less susceptible to corruption by noise, etc. The notch is centered towards the upper end of the line characteristic plateau to further improve transmission. A description will now be given of system operation:

A subscriber uses telephone 1 to make a telephone connection with the operator telephone 5 through the public network 4. The telephone channel provides good quality full duplex speech. An operator or voiceographer at the operator station (5, 6) then receives dictation from the subscriber telephone 1 and- types a voice message onto the keyboard 6c of computer 6a. This loads the voice message, in digital form, into a buffer (not shown) in computer 6a where it is stored and used by wordprocessing software to illustrate the text on the display 6b.

As each key is struck by the operator, a scan code is generated and this in turn generates the appropriate ASCII code from a translate table. The ASCII code is then split into 4 four bit portions and a hamming code is added to each portion to make up 8 bits. The 8 bit portions are then duplicated and the resultant data segment is prefixed with a duplicate two byte hexadecimal AA. This data segment is then stored in the computer buffer.

Thus, the software of computer 6a traps the hardware interrupts caused by each key stroke so that each time a key is operated, the software:

(i) takes control of the computer 6a; (ii) allows for the hardware key code to be translated to ASCII by whatever standard software is used (e« « MS-DOS program KEY B) ; and

(iii) copies the ASCII code from the keystroke buffer, encodes it into a short sequence of binary digits or bits by a special technique which includes the use of "Hamming Codes" for Forward Error Correction, and then transmits the encoded sequence of bits, via the RS 232C serial port to the data modem 7d, filter 7c, and line interface 7b to the public network.

Whilst the modulated carrier is passed to the line interface 7b by the data modem 7d, the text generated by the keyboard 6a is displayed on the operator's display screen 6b so that the operator can verify that the voice message has been correctly typed. This process does not interfere with voice communication since voice signals pass the notch filter 3b before transmission on the PSTN 4 and these signals then enter the line interface 7b before passing the complementary notch filter 7a for reception on the operator's telephone 5. Filter 7c prevents these voice signals from interfering with the output of data modem 7b.

The data signals received at the subscriber station pass the line interface 3d and filter 3c before they are demodulated by the data modem 3b. They then enter computer 2a via the RS 232C serial port where the software of computer 2a traps the hardware interrupts caused by the arrival of the data.

Each time an encoded sequence of bits, representing a key stroke at the operator's computer 6a, arrives at the subscriber computer 2a, the software:

(i) takes control of computer 2a;

(ii) decodes the sequence of bits which arrive in order to recreate the original ASCII code for the key which the operator pressed; and (iϋ) places the ASCII code in the subscriber computer's key stroke buffer just as if a key has been pressed by the subscriber on his own keyboard 2b.

One way of viewing the operation is to consider computer 6a as a master and computer 2a as a slave. In this way, the master computer 6a remotely controls the slave computer

2a during the transmission of data.

After the ASCII key code generated by computer 6a has been placed in the keystroke buffer of the slave computer 2a, the software effectively relinquishes control of the slave computer 2a. Computer 2a allows the ASCII key code to be sent to the application program (wordprocessor) as in normal applications. The arrangement is such that it is not apparent to the application program that the software of has intervened and the application program of computer 2a uses the keystrokes as normal command/data entries.

The master slave concept can be used in reverse in order to reverse the direction of dictation and the entry of a voice message on a keyboard.

The communications system can also be used to provide real time translation when a conference network has been established over a PSTN or a private communication link. In this way, the operator is a translator who receives either a voice message in a foreign language, or text in a foreign language with data modem 7b in a "data receive" mode and then translates the voice message or text so that it can be transmitted back to the subscriber. Ultimate error correction is by human intervention since the subscriber can view the text on his display screen 2b, character by character and, should any error occur, e.g. due to misunderstanding, erroneous keystrokes or corruption due to severe line noise which corrupts the keystroke information transmitted from the master to the slave computer, then the subscriber can request an immediate correction via the duplex voice channel which runs continuously. The subscriber may also instruct the operator to punctuate, embolden, underscore, delete or otherwise edit the text at any time.

The created file becomes resident in both the master and slave computer and the subscriber can further process the document using his own secretarial services.

Once the. document has been completed, it may be transmitted to the. another computer, any electronic mail box, any telex terminal, or any facsimile machine in the world.

The general process described above may also be provided at a local level within an office or the same building. In this case, speech communication is established via the subscriber telephone 1 through a direct line or PABX 8 to the operator's telephone 5 as shown in Fig. 3. The difference between this local system and the more general system shown in Fig. 1 is that the computers 2, 6 are connected by a separate data channel 9 which is independent of the voice channel 10 connecting the telephones 1, 5. The data channel 9 may optionally include a line driver 11 (depending on the distance between the computers 2, 6) and a switch 12 for connecting different operator terminal to the same data channel.

A further embodiment of the invention is illustrated in Fig. 4 where a subscriber initiates a communication link through a digital telephone la connected to a fifth generation 2B+D PABX 13. After switching to an appropriate extension port, digitised speech is transmitted via a 64 kB digitised voice channel 13 to the operator's digital telephone 5b to provide high quality full duplex speech. The line connecting the 2B+D PABX 13 to each telephone also provides a 64 kB data channel and a control channel. As before, the subscriber dictates a voice message and the operator generates keyboard signals which are transmitted over the communication link for display at the subscriber station.

Yet a further alternative (not illustrated) is equivalent to Fig. 4. However, an Integrated Switched Data Network (ISDN) is employed instead of the 2B+D PABX 13. Each Integrated Data Access (IDA) telephone line provides one 64 kB digital speech channel, one 64 kB data channel and a control channel. Also, IDA telephones replace the digital telephones la and 5b shown in Fig. 4. The IDA telephones each provide an integral RS 232C serial data interface port.

Modifications and improvements are possible without departing from the scope of the invention and the foregoing embodiments have been described only by way of example.

Claims

1. A method of communication between a subscriber station and an operator station characterised by the steps of: a) dictating a voice message from the subscriber station to the operator station over a voice channel, b) generating keyboard signals at the operator station corresponding with characters in text representing the voice message, and c) causing respective computers at each station to operate so that data representing the keyboard signals is transmitted from the operator station to the subscriber station over a data channel to enable the keyboard signals to be displayed at the subscriber station.

2. A method according to claim 1 in which step (b) includes encoding the keyboard signals with special codes which serve for recognition and for error correction at the subscriber station.

3. A method according to claim 1 or 2 in which step (a) includes notch filtering signals representing the voice message and step (c) includes filtering the data for transmission in the notch.

4. A communications system comprising: a subscriber station and an operator station; a respective computer at each station, each computer including a display for viewing text created by digitised keyboard signals stored in a respective buffer, a keyboard being provided at least at the operator station; a data channel enabling data to be exchanged between the computers; a voice channel enabling voice communication between the stations; either said data and said voice channels being independent, or means being provided to enable the data to be exchanged and the voice communication to occur during the same call; characterised in that said computers are programmed to cause the system to operate so that each keyboard signal is encoded at the operator station in a manner in which it can be decoded and recognized as a keyboard signal at the subscriber station whereby, as the subscriber dictates a voice message over the voice channel, data representing each character keyed by the operator at the operator station is transmitted over the data channel and decoded so that it can be viewed on the display at the subscriber station.

5. A system according to claim 4 wherein the programming is such that when each key is struck on the operator's keyboard, it causes an appropriate scan code and ASCII code to be generated, the interrupt due to the keystroke being used to trigger a sub-routine to split the encoded and digitised signal into parts which are further encoded with Hamming codes to ensure error-free decoding after transmission; the resultant data segment being prefixed with a recognition code to which the computer at the subscriber station responds as if it were a locally generated keystroke interrupt and the computer at the subscriber station decoding the received data so as to reproduce a digitised signal which is stored in the buffer as a standard character signal for display.

6. A system according to claim 5 in which coded portions of both the data segment and the prefix are doubled, prior to transmission, to ensure receipt of at least one portion that is not corrupted by noise.

7. A system according to any of claims 4-6 in which both the voice and the data channels share the same communications link, filters being provided so as to create a notch in the voice band of the voice channel, the notch enabling the data channel to operate simultaneously within the voice channel.

8. A system according to any of claims 4-6 in which, the voice and data channels are independent of one another.

9. A system according to any of claims 4-6 in which the voice and data channels are provided by a digital telephone connected to a fifth generation 2B+D PABX having an RS 232C serial port for entering a data stream, by an Integrated Switched Data Network (ISDN) having an integral RS 232C serial data interface port.