GB2357681A

GB2357681A - Inband signalling within speech segments

Info

Publication number: GB2357681A
Application number: GB9930571A
Authority: GB
Inventors: David Fraser Chambers; Stephen Basil Aftelak
Original assignee: Motorola Ltd
Current assignee: Motorola Solutions UK Ltd
Priority date: 1999-12-23
Filing date: 1999-12-23
Publication date: 2001-06-27
Also published as: WO2001048735A2; WO2001048735A3; AU1392501A; GB9930571D0

Abstract

A method and apparatus for speech transmission including inband signalling, comprises a sampler 3 for digitally sampling the speech signal to product blocks of bits representing the speech signal, means 15 for providing a queue of inband signals to be transmitted with the speech signal, a detector 13 for monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of the inband signals, and an insertion device 9 for inserting inband signals from the queue 15 into the blocks of bits representing the speech signal during the periods of speech with low likelihood of degradation. As a result, inband signals can be transmitted with a speech signal without degrading the apparent quality of the speech signal for a user of the system. The inband signals are preferably inserted into the least significant bits of the blocks, either in bursts of consecutive blocks or in 1 in N blocks.

Description

1 2351681 INBAND SIGNALLING WITHIN SPEECH SEGMENTS

Field of the Invention

The present invention relates to communications systems, and in particular to inband signalling within speech segments being transmitted.

Background of the Invention

When communicating, speech may be transmitted from person to person via land lines, via satellite or, if at least one of the people is using a mobile telephone, via a radio frequency signal. The present invention relates to such transmissions which are of a digital nature.

In telecommunications, pulse code modulation (PCM) is an accepted method for sampling and encoding analogue speech from a microphone, to produce, say, a series of eight bit blocks for transmission. The PCM used by telecommunications organisations is defined by standard G.711 produced by the International Telecommunications Union (ITU). Hence, speech is usually transmitted, in digital applications, as encoded eight bit blocks. In the U.S., the encoding is usually achieved using p-law encoding, whereas in Europe the encoding is usually done using A-law encoding.

When transmitting a speech signal using PCM, it is often necessary to transmit 25 related information to assist the receiving end of the transmission to understand how the transmission has been encoded, the speed at which the encoding has taken place, etc. Hence, it is sometimes necessary to include inband signalling within the speech segments of a transmission.

To provide space for the inband signalling, it is common to "steal" resources from the underlying PCM samples. For example, a scheme may steal the least significant bit (LSB) of M in W PCM samples or, alternatively, the LSB from a defined number of consecutive PCM samples. The least significant bit (LSB) is 2 the bit which will be missed least if it is not used for the speech signal transmission.

More particularly, inband signalling can be implemented by periodically stealing the least significant bit (LSB) of PCM speech samples. Such stealing can also be implemented on a burst basis by permanently stealing the least significant bit for fixed periods of time interspersed with periods where none of the LS13's are stolen. This burst-mode technique has advantages over techniques which derive a signalling channel by stealing M in W LS13's on a semi-permanent basis. Firstly, a given amount of signalling data can be transferred in a much shorter time (i.e. a factor of N, where N may be as large as 16). Secondly, identification and synchronisation to the signalling message can be much simplified, since synchronisation to the "1 in W pattern is not required in addition to synchronisation to the signalling message itself. Conversely, burst-mode signalling can have a greater detrimental effect on speech quality than the M in W approach.

Indeed, both techniques can lead to undesirable artefacts which are audible in the speech signal. During quiet periods in particular, bit stealing seems to affect speech quality. Further, if the burst-mode method is used, periods of slightly poorer speech transmission quality may be interspersed with periods of good speech transmission quality, which could also become annoying to a user over the length of a complete call. With this in mind, intermittent interference may be more annoying than a speech transmission which has less than perfect quality throughout.

In view of the foregoing, there is a clear need to improve upon the effect of inband signalling on speech transmissions, and the present invention aims to satisfy this need.

3 Summary of the Invention

The aim of the present invention is to reduce or even eliminate audible effects due to inband signalling. More particularly, the present invention aims to use the 5 inherent masking capabilities of the speech signal itself to "hide" inband signalling.

With the foregoing in mind, the present invention provides a method of speech transmission including inband signalling, comprising: digitally sampling a speech signal to produce blocks of bits representing the speech signal; providing a queue of inband signals to be transmitted with the speech signal; monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of inband signals; and inserting inband signals from the queue into the blocks of bits representing the speech signal during the periods of speech with low likelihood of degradation.

By monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of inband signals, the inband signals can be inserted at times of least detrimental effect. This is something which provides a distinct advantage over previous methods of speech transmission including inband signalling.

Although it is known from M.R. Schroeder, B.S. Atal and J.L. Hall, "Objective Measure of Certain Speech Signal Degradations Based on Masking Properties of Human Auditory Perception", Frontiers of Speech Communication, ed. Lindblom & Ohman, Academic Press, 1979, that speech degradations can be masked by the speech signal itself, a method of monitoring a speech signal for periods of speech with the likelihood of degradation is completely new in the field of speech transmission including inband signalling.

In a particular embodiment the inband signals may be inserted only into least significant bits of the blocks of bits representing the speech signal. By doing this, 4 the quality of the speech transmission will be even less deteriorated by the insertion of the inband signals.

In one embodiment, the least significant bits may be in bursts of consecutive blocks. In another embodiment, the least significant bits of M in M' blocks may be used during the periods of speech with low likelihood of degradation. Typically, the periods of speech with low likelihood of degradation may be of the order of 20 milliseconds.

The method may include delaying the sampled speech signal whilst the monitoring takes place. This is to ensure that the inband signals are inserted during the periods of speech with low likelihood of degradation, which will only be known once the speech signal has been monitored.

Once the signal for transmission has been prepared, it may be transmitted via a radio frequency signal, by cable or land line, or any other appropriate means.

As will be appreciated, the method of speech transmission preferably includes encoding the sample speech signal prior to transmission.

The present invention further provides an apparatus for inserting inband signalling into a speech signal for onward transmission, comprising:

a sampler for digitally sampling a speech signal to produce blocks of bits representing the speech signal; input means for receiving a queue of inband signals to be transmitted with the speech signal; a detector for monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of inband signals; and an insertion device for inserting inband signals from the queue into the blocks of bits representing the speech signal.

Detailed Description of the Drawings

A specific embodiment of the present invention will now described, by way of example only, with reference to the accompanying drawing which shows a schematic representation of a circuit defining the method and apparatus of the present invention.

With reference to the drawing, there is shown an input 1 for the speech signals. The main route for the speech signals is through a sampler 3, an encoder 5, a delay 7 and an insertion device 9 to the output 11. The speech signals are also monitored by a detector 13 which searches for periods of speech with low likelihood of degradation if inband signals are inserted into the sampled and encoded speech signals. A series of inband signals also enter the apparatus and are held in a queue 15 by some form of switch 17.

When the detector 13 notes a period of speech with low likelihood of degradation, it produces an output which results in closing of the switch 17. An inband signal in the queue 15 is then able to travel to the insertion device 9 where it is inserted into the sampled and encoded speech signal for onward transmission via the output 11. The delay 7 is used to ensure that an inband signal which is to be added to the speech signal is added at the correct time.

Although not shown specifically in the drawing, the sampling and encoding of the speech signal is done using pulse code modulation and the inband signals are inserted in place of the least significant bits of at least some of the blocks of bits produced as a result of the digital sampling of the speech signals.

If the periods of speech with low likelihood of degradation due to the insertion of inband signals are short, the inband signals may be broken up into smaller components for onward transmission during the periods of speech with low likelihood of degradation. With this in mind, however, typical periods of speech with low likelihood of degradation are of the order of 20 to 90 milliseconds in length 6 and this would normally be sufficiently long to transmit a useful inband signalling message.

It will of course be understood that the present invention has been described above purely by way of example, and that modifications of detail can be made within the scope of the appendant claims.

7

Claims

1 A method of speech transmission including inband signalling, comprising: digitally sampling a speech signal to produce blocks of bits representing the speech signal; providing a queue of inband signals to be transmitted with the speech signal; monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of inband signals; and inserting inband signals from the queue into the blocks of bits representing the speech signal during the periods of speech with low likelihood of degradation.

2. A method as claimed in claim 1, wherein the inband signals are inserted into least significant bits of the blocks of bits representing the speech signal.

3. A method as claimed in claim 2, wherein the least significant bits are in bursts of consecutive blocks.

4. A method as claimed in claim 2, wherein the least significant bits of 1 in N blocks during the period of speech with low likelihood of degradation are used.

5. A method as claimed in any preceding claim, further comprising: delaying the sampled speech signal whilst the monitoring occurs.

6. A method as claimed in any preceding claim, further comprising:

transmitting the blocks of bits via a radio frequency signal or by cable.

7. A method as claimed in any preceding claim, further comprising:

encoding the sampled speech signal prior to transmission.

8. A method of speech transmission including inband signalling substantially as hereinbefore described with reference to and as shown in the accompanying drawing.

1 8

9. An apparatus for inserting inband signalling into a speech signal for onward transmission, comprising: a sampler (3) for digitally sampling a speech signal to produce blocks of bits representing the speech signal; an input for receiving a queue (15) of inband signals to be transmitted with the speech signal; a detector (13) for monitoring the speech signal for periods of speech with low likelihood of degradation due to the insertion of inband signals; and an insertion device (9) for inserting inband signals from the queue (15) into the blocks of bits representing the speech signal during the periods of speech with low likelihood of degradation.

10. An apparatus as claimed in claim 9, further comprising:

a transmitter for transmitting the sampled speech signal with inband signals.

11. An apparatus as claimed in claim 9 or claim 10, further comprising a delay (7) for delaying the sampled speech signal whilst the detector (13) monitors the speech signal for periods of speech with low likelihood of degradation.

12. An apparatus as claimed in any one of claims 9 to 11, further comprising: an encoder (5) for encoding the digitally sampled speech signal prior to transmission.

13. An apparatus as claimed in any one of claims 9 to 12, further comprising a switch (17) which is activated by the detector (13) to control the queue (15) of inband signals.

14. An apparatus substantially as hereinbefore described with reference to and as shown in the accompanying drawing.