Rudimentary digital speech interpolation apparatus and method

• US 5,065,395 A
• Filed: 04/09/1990
• Issued: 11/12/1991
• Est. Priority Date: 04/09/1990
• Status: Expired due to Term

• Current Assignee: Alcatel-Lucent USA, Inc. (Nokia Corporation)
• Original Assignee: DSC Communications Corp. (Nokia Corporation)
• Inventors: Shenoi, Kishan, Yu, Frank I., Ho, Helena S., Hanagan, Patrick L.
• Primary Examiner(s): Olms, Douglas W.
• Assistant Examiner(s): KIZOU, HASSAN
• Application Number:
  US07/506,678
• Time in Patent Office: 582 Days
• Field of Search: 370/81, 370/79, 370/80, 370/118, 381/46
• US Class Current: 370/435
• CPC Class Codes: H04J 3/172 Digital speech interpolation, i.e. DSI

Abstract

Rudimentary digital speech interpolation apparatus for compressing data on a plurality of channels is disclosed, which includes circuitry for receiving data bits and control bits for each channel. The presence of silence on each channel is determined in response to the received data bits and zero bits are allocated for any silent channel in a frame structure. In addition, circuitry also allocates x bits per channel for the data bits in an non-silent channel in the frame. At least six bits per channel are also allocated for the control bits in the frame. 4N such frames are grouped to form a multi-frame.

Claims

1. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining whether each channel is silent or non-silent at time t by determining whether the signal on each channel has been below a threshold level since (t-T), where T is a predetermined time period;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels; and

means for forming a multi-frame having a plurality of said frames.

2. The apparatus, as set forth in claim 1, wherein a channel is deemed silent if said channel has been below said predetermined threshold since time (t-100 msec).

3. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent, said means for determining comprising;

a bandpass filter for eliminating zero frequency components and 4 KHz components of a signal present on each channel;

a squarer for receiving said filtered signal for squaring the filtered signal and producing a squared signal;

a lowpass filter for receiving said squared signal and eliminating high frequency components in said squared signal and producing a short-term power of said signal on each channel;

a comparator for comparing said short-term power with a predetermined silence threshold;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels; and

means for forming a multi-frame having a plurality of said frames.

4. The apparatus, as set forth in claim 3, wherein said predetermined silence threshold is approximately -45dBm0.

5. The apparatus, as set forth in claim 3, wherein said predetermined silence threshold is approximately 3 dB above a background noise level present on each channel.

6. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels;

means for forming a multi-frame having a plurality of said frames; and

wherein said silence determining means comprises;

a bandpass filter for eliminating zero frequency components and 4 KHz components of a signal present on each channel and producing a sequence w(n);

wherein said silence determining means further computes ##EQU2## to produce z(n), a short-term power measurement of said signal on each channel; and

a comparator for comparing said short-term power measurement with a predetermined silence threshold.

7. The apparatus, as set forth in claim 6, wherein said predetermined silence threshold is approximately -45dBm0.

8. The apparatus, as set forth in claim 6, wherein said predetermined silence threshold is approximately 3 dB above a background noise level on each channel.

9. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels;

means for forming a multi-frame having a plurality of said frames; and

wherein said data on said plurality of channels are compressed by waveform encoding techniques to x bits per sample, where x≦

8, and said control bits include the number of bits per sample x for each channel for the duration of a multi-frame.

10. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent;

means for producing a plurality of control bits for each of said plurality of channels, said control bits for each of said plurality of channels include background noise level for said channel;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels; and

means for forming a multi-frame having a plurality of said frames.

11. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels;

means for forming a multi-frame having a plurality of said frames; and

wherein said control bits for each of said plurality of channels further include a plurality of A/B signaling bits.

12. Apparatus for compressing data on a plurality of channels by digital speech interpolation, comprising:

means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether each channel is silent or non-silent;

means for producing a plurality of control bits for each of said plurality of channels;

means for forming a plurality of frames each having data of said non-silent channels and a plurality of control bits of one of said plurality of channels;

means for forming a multi-frame having a plurality of said frames; and

wherein said control bits for each of said plurality of channels further include system status information such as alarms.

13. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining whether a channel is silent or non-silent at time t by determining whether each channel has been below a threshold level since time (t-T), where T is a predetermined time period;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame; and

means for grouping a plurality of said frames to form a multi-frame.

14. The apparatus, as set forth in claim 13, wherein a channel is deemed silent if said channel has been below said predetermined threshold since time (t-100 msec).

15. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent, said means coupled comprising;

a bandpass filter for eliminating zero frequency components and 4 KHz components of a signal present on each channel;

a squarer for receiving said filtered signal for squaring the filter signal and producing a squared signal;

a lowpass filter for receiving said squared signal and eliminating high frequency components in said squared signal and producing a short-term power of said signal on each channel; and

a comparator for comparing said short-term power with a predetermined silence threshold;

means for forming a frame;

means for allocating zero bits in said frame for each silent chamber;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame; and

means for grouping a plurality of said frames to form a multi-frame.

16. The apparatus, as set forth in claim 15, wherein said predetermined silence threshold is approximately -45dBm0.

17. The apparatus, as set forth in claim 15, wherein said predetermined silence threshold is approximately 3 dB above a background noise level present on each channel.

18. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame;

means for grouping a plurality of said frames to form a multi-frame; and

wherein said silence determining means comprises;

a bandpass filter for eliminating zero frequency components and 4 KHz components of a signal present on each channel and producing a sequence w(n);

wherein said silence determining means further computes ##EQU3## to produce z(n), a short-term power measurement of said signal on each channel; and

a comparator for comparing said short-term power measurement with a predetermined silence threshold.

19. The apparatus, as set forth in claim 18, wherein said predetermined silence threshold is approximately -45dBm0.

20. The apparatus, as set forth in claim 18, wherein said predetermined silence threshold is approximately 3 dB above a background noise level on each channel.

21. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating x bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame;

means for grouping a plurality of said frames to form a multi-frame; and

wherein said data on said plurality of channels are compressed by waveform encoding techniques to x bits per sample, where x≦

8, and said control bits include the number of bits per sample x for each channel for the duration of a multi-frame.

22. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame, said control bits for each of said plurality of channels include background noise level for said channel; and

means for grouping a plurality of said frames to form a multi-frame.

23. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame;

means for grouping a plurality of said frames to form a multi-frame; and

wherein said control bits for each of said plurality of channels further include a plurality of A/B signaling bits.

24. Digital speech interpolation apparatus for compressing data on a plurality of channels, comprising:

means for receiving data bits and control bits for each channel;

means coupled to said receiving means for determining the presence of silence on each of said plurality of channels, and determining therefrom whether a channel is silent or non-silent;

means for forming a frame;

means for allocating zero bits in said frame for each silent channel;

means for allocating a predetermined number of bits in said frame for each non-silent channel for accommodating said data bits;

means for allocating a plurality of bits for one of said plurality of channels for accommodating said control bits in said frame;

means for grouping a plurality of said frames to form a multi-frame; and

wherein said control bits for each of said plurality of channels further include system status information such as alarms.

25. A digital speech interpolation method used in a data compression system, said data compression system compressing data on 4N channels with waveform encoding techniques and producing encoded data bits and control bits, said method comprising the steps of:

receiving said encoded data bits and control bits for said 4N channels;

determining whether each channel is below a predetermined threshold at time t;

determining whether a channel which is below said threshold at time t has been below said threshold for a predetermined time period, so that such channels are deemed silent;

allocating zero bits per sample for each silent channel;

grouping every M encoded data bis and control bits to form a frame, said data bits in each frame including encoded data bits for all of 4N channels which are active and not silent; and

grouping 4N frames to form a multi-frame.

26. The method, as set forth in claim 25, wherein said step of determining silent channels determines whether a channel has been silent for 100 msec prior to time t.

27. The method, as set forth in claim 25, further comprising the step of determining background noise level in each channel.

29. The method, as set forth in claim 25, wherein said step of determining whether a channel is below the predetermined threshold further comprises:

filtering a signal present on said channel to eliminate zero frequency components and 4 KHz components;

squaring the filtered signal and producing a squared signal;

eliminating high frequency components in said squared signal and producing a short-term power of said signal on said channel; and

comparing said short-term power with a predetermined silence threshold.

30. The method, as set forth in claim 29, wherein said step of comparing compares said short-term power with a predetermined silence threshold of approximately -45dBm0.

31. The method, as set forth in claim 29, wherein said step of comparing compares said short-term power with a predetermined silence threshold of approximately 3 dB above a background noise level.

28. A digital speech interpolation method use din a data compression system, said data compression system compressing data on 4N channels with waveform encoding techniques and producing encoded data bits and control bits, said method comprising the steps of:

receiving said encoded data bits and control bits for said 4N channels;

determining whether each channel is below a predetermined threshold level at time t;

determining whether a channel below the predetermined threshold level has been below said predetermined threshold level since time (t-T), where T is a predetermined time period, so that such channels are deemed silent;

determining background noise level in each channel;

allocating zero bits per sample for each silent channel;

grouping every M encoded data bits and control bits to form a frame, said data bits in each frame including encoded data bits for all of 4N channels which are active and not silent;

grouping 4N frames to form a multi-frame; and

wherein said control bits in frame n of a multi-frame include the number of encoded bits per sample for channel n for the duration of the multi-frame, and the background noise level of channel n for the duration of the multi-frame.

32. A rudimentary digital speech interpolation method use in a data compression system, said data compression system comprising data on 4N channels with waveform encoding techniques and producing encoded data bits and control bits, said method comprising the steps of:

receiving said encoded data bits and control bits for said 4N channels;

determining whether each channel is below a predetermined threshold at time t;

determining whether a channel below the predetermined threshold at time t has been below the predetermined threshold since time (t-T), where T is predetermined time period, so that such channels are deemed silent;

allocating zero bits per sample for each silent channel;

grouping every M encoded data bits and control bits to form a frame, said data bits in each frame including encoded data bits for all of 4N channels which are active and not silent;

grouping 4N frames to form a multi-frame; and

wherein said step of determining whether a signal on a channel is below a predetermined threshold, comprises the steps of;

filtering to eliminate zero frequency components and 4 KHz components of said signal and producing a sequence w(n);

computing ##EQU4## to produce z(n), a short-term power measurement of said signal in a channel; and

comparing said short-term power measurement with a predetermined silence threshold.

33. The method, as set forth in claim 32, wherein said comparing step compares said short-term power measurement with said predetermined silence threshold of approximately -45dBm0.

34. The method, as set forth in claim 32, wherein said comparing step compares said short-term power measurement with said predetermined silence threshold of approximately 3 dB above a background noise level present on each channel.