Method and system for data transmission using a lossy compression service

• US 6,327,303 B1
• Filed: 07/30/1998
• Issued: 12/04/2001
• Est. Priority Date: 07/30/1998
• Status: Expired due to Term

• Current Assignee: Google Technology Holdings LLC (Alphabet Inc.)
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: Balogh, Craig Robert, Yip, William Chun-Hung, Hall, Timothy Gerard
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): Ha, Dac V.
• Application Number:
  US09/124,785
• Time in Patent Office: 1,223 Days
• Field of Search: 375/240, 375/240.11, 375/240.14, 375/240.15, 375/240.25, 375/240.26, 704/500, 704/503
• US Class Current: 375/240
• CPC Class Codes: H04B 1/66 for reducing bandwidth of signals; for improving efficiency of transmission H04B1/68 takes precedence

Abstract

A system (20) transmits a digital data stream (22) using a lossy compression service (24) having a communication path (28) through which compressed signals propagate. The system includes a data conditioner (36) configured to condition the received digital data stream (22) to produce data subsets (66). A frame generator (36) expands each of the data subsets (66) to generate distinct signal sets (68). The distinct signal sets (68) are expressed in a data size greater than the data subsets (66) to compensate for lossy compression cause by the lossy compression service (24). A transmitter is in communication with the frame generator (36) and transmits digitized ones of the distinct signal sets (68) in frames of a digital bit stream (126) over the communication path (28).

Claims

1. A system for transmitting a digital data stream using a lossy compression service having a communication path through which compressed signals propagate, said system comprising:

a data conditioner configured to receive said digital data stream, said data conditioner conditioning said digital data stream to produce a plurality of data subsets;

a frame generator in communication with said data conditioner for expanding each of said data subsets to generate distinct signal sets, said distinct signal sets being expressed in a data size greater than said data subsets to compensate for lossy compression caused by said lossy compression service; and

a transmitter in communication with said frame generator for transmitting digitized ones of said distinct signal sets in frames of a digital bit stream over said communication path.

2. A system as claimed in claim 1 wherein said frame generator comprises a table for associating data subset entries with said distinct signal sets, said data subset entries being indexed in response to each of said data subsets to obtain said distinct signal sets.

3. A system as claimed in claim 2 wherein said table comprises:

a plurality of memory locations, each of said memory locations being configured to store one of said data subset entries representative of one of said data subsets; and

a plurality of separate signal-defining patterns, one each of said separate signal-defining patterns being stored in association with one each of said data subset entries, wherein one each of said separate signal-defining patterns represents one each of said distinct signal sets.

4. A system as claimed in claim 3 wherein said frame generator further comprises a waveform generator for receiving said separate signal-defining patterns, said waveform generator being configured to generate said distinct signal sets in response to receipt of said separate signal-defining patterns from said table.

5. A system as claimed in claim 4 wherein said lossy compression service is a digital voice service, said transmitter is a vocoder encoder system, said compressed signals are vocoder signals which propagate over said communication path from said vocoder encoder system to a vocoder decoder system of said digital voice service, said distinct signal sets are distinct digitized audio waveforms, each of said separate signal-defining patterns describes a separate audio spectrum for each of said distinct digitized audio waveforms, and said table is configured to associate one each of said separate spectra with one each of said data subset entries.

6. A system as claimed in claim 5 wherein:

said data conditioner is configured to condition said digital data stream to produce said plurality of data subsets each having a bit number compatible with a predetermined number of bits in each of said frames of said digital bit stream;

said waveform generator is configured to generate said distinct audio waveforms in response to receipt of said ones of said separate spectra from said table; and

said vocoder encoder is configured to compress said distinct audio waveforms to said bit number for transmission in said frames of said digital bit stream.

7. A system as claimed in claim 1 further comprising:

a receiver for receiving said digital bit stream from said communication path, said receiver generating synthesized ones of said distinct signal sets;

a frame detector in communication with said receiver for transforming said synthesized distinct signal sets to form synthesized signal-defining patterns, each of said synthesized signal-defining patterns being associated with synthesized ones of said plurality of data subsets; and

a data assembler in communication with said frame detector and configured to receive said synthesized ones of said data subsets in response to said synthesized signal-defining patterns, said data assembler assembling said synthesized ones of said data subsets to produce a synthesized digital data stream representative of said digital data stream.

8. A system as claimed in claim 7 wherein said frame detector comprises a table for storing said separate signal-defining patterns in memory locations of said table relative to said synthesized ones of said data subsets, and said frame detector is configured to select ones of said separate signal-defining patterns that most closely match said synthesized signal-defining patterns in order to obtain said synthesized data subsets.

9. A system as claimed in claim 7 wherein:

said frame generator is configured to add frame markers to said generated distinct signal sets to synchronize said generated distinct signal sets relative to temporal frames; and

said frame detector is configured to detect said frame markers following transmission of said digital bit stream over said communication path.

10. A system as claimed in claim 7 wherein said frame detector comprises a spectrum generator for generating said synthesized ones of said separate signal-defining patterns in response to said synthesized distinct signal sets.

11. A method for transmitting a digital data stream using a lossy compression service having a communication path through which compressed signals propagate, said method comprising the steps of:

relating, in a first table, a distinct signal set to each of a first plurality of data subsets;

conditioning said digital data stream to produce a second plurality of data subsets, each of said second plurality of data subsets being any one of said first plurality of data subsets;

indexing ones of said first plurality of data subsets in said first table in response to each of said second plurality of data subsets to obtain related ones of said distinct signal sets;

transmitting a digital bit stream representative of said related distinct signal sets over said communication path of said lossy compression service;

accessing a second table in response to received frames of said digital bit stream to synthesize said second plurality of data subsets of said digital data stream; and

assembling said synthesized second plurality of data subsets to synthesize said digital data stream.

12. A method as claimed in claim 11 wherein said relating step comprises the step of expanding each of said first plurality of data subsets to generate each of said distinct signal sets such that said distinct signal sets are expressed in a size greater than said data subsets in order to compensate for lossy compression caused by said lossy compression service.

13. A method as claimed in claim 11 wherein said relating step occurs in a table initialization process which is performed prior to said conditioning step.

14. A method as claimed in claim 11 wherein each data subset of said first plurality of data subsets is a fixed-size bit stream exhibiting a unique bit pattern, said first plurality of data subsets is configured to include every combination of said unique bit patterns, and said relating step comprises the steps of:

identifying a plurality of distinct signal sets compatible with transmission through said communication path;

assigning one each of said plurality of distinct signal sets to one each of said first plurality of data subsets; and

associating each of said first plurality of data subsets with said assigned ones of said distinct signal sets in said first table.

15. A method as claimed in claim 14 wherein said lossy compression service is configured to process frames of lossy-compressed digital bit patterns, and said identifying step comprises the steps of:

decoding said frames of lossy-compressed digital bit patterns to produce signal sets;

transforming said signal sets to obtain a separate signal-defining pattern for each of said signal sets;

measuring a difference between each of said separate signal-defining patterns and a preselected set of separate signal-defining patterns to obtain an error value for each of said separate signal-defining patterns;

distinguishing ones of said separate signal-defining patterns in which said measured total error values are greater than a predetermined error threshold; and

adding said distinguished ones of said separate signal-defining patterns to said preselected set of separate signal-defining patterns, said preselected set of separate signal-defining patterns representing said plurality of distinct signal sets.

16. A method as claimed in claim 15 wherein said measuring step includes the step of calculating a Euclidean distance measure for each of said separate signal-defining patterns to produce said total error value.

17. A method as claimed in claim 15 wherein said lossy compression service is a digital voice service, said compressed signals are vocoder signals which propagate over said communication path from a vocoder encoder to a vocoder decoder of said digital voice service, said distinct signal sets are distinct digitized audio waveforms compatible with said digital voice service, and each of said separate signal-defining patterns is a separate spectrum.

18. A method as claimed in claim 11 wherein said digital bit stream includes frames of lossy-compressed digital bit patterns, each of said frames having a predetermined number of bits, and said conditioning step comprises the step of parsing said digital data stream into data subsets selected from said second plurality of data subsets, each of said data subsets containing a quantity of bits less than said predetermined number of bits in each of said frames.

19. A method as claimed in claim 11 wherein:

said conditioning step comprises the step of executing a linear transformation function on said digital data stream to form said second plurality of data subsets prior to said indexing step; and

said assembling step comprises the step of executing an inverse linear transformation function on said synthesized second plurality of data subsets to form said synthesized digital data stream.

20. A method as claimed in claim 11 wherein:

said indexing step includes the step of obtaining separate signal-defining patterns of said related ones of said distinct signal sets; and

said method further comprises the step of transforming said separate signal-defining patterns to generate said distinct signal sets prior to said transmitting step.

21. A method as claimed in claim 11 wherein said lossy compression service converts each of said distinct signal sets into frames of a predetermined number of bits, and said method further comprises the step of adding a frame marker to each of said distinct signal sets prior to said transmitting step.

22. A method as claimed in claim 21 further comprising the steps of:

receiving said frames of said digital bit stream at a receiver of said lossy compression service;

converting said digital bit streams into synthesized ones of said distinct signal sets;

detecting said frame marker; and

transforming said synthesized distinct signal sets in response to said detected frame marker prior to said accessing step to obtain synthesized separate signal-defining patterns of said synthesized distinct signal sets.

23. A method as claimed in claim 22 wherein said distinct signal sets are represented as separate signal-defining patterns in said second table, and said accessing step comprises the steps of:

selecting one of said separate signal-defining patterns from said second table that most closely matches one of said synthesized separate signal-defining patterns; and

obtaining one of said data subsets from said first plurality of data subsets related to said selected separate signal-defining pattern.

24. A method for transmitting a digital data stream using a digital wireless service having a communication path through which vocoder signals propagate, said method comprising the steps of:

relating, in a first table, a distinct audio waveform to each of a first plurality of data subsets;

conditioning said digital data stream to produce a second plurality of data subsets, each of said second plurality of data subsets being any one of said first plurality of data subsets;

indexing ones of said first plurality of data subsets in said first table in response to each of said second plurality of data subsets to obtain related ones of said distinct audio waveforms;

generating said distinct audio waveforms in response to said indexing step;

adding a frame marker to each of said generated distinct audio waveforms;

encoding, at an encoder system of said digital wireless service, each of said distinct audio waveforms to obtain frames of a vocoder digital bit stream;

transmitting said vocoder digital bit stream representative of said related distinct audio waveforms over said communication path of said digital wireless service;

decoding, at a decoder system of said digital wireless service, said vocoder digital bit stream to generate synthesized distinct audio waveforms;

detecting said frame markers to differentiate each of said synthesized distinct waveforms;

transforming each of said synthesized distinct audio waveforms to produce synthesized spectra;

accessing a second table to select a most closely matching spectrum for each of said synthesized spectra to obtain related ones of said second plurality of data subsets; and

assembling said second plurality of data subsets to synthesize said digital data stream.

25. A method as claimed in claim 24 wherein the digital data stream comprises video data.

26. In a system for transmitting a digital data stream using a digital lossy compression service, a method of training a table to relate data subsets of said digital data stream with distinct signal sets, said method comprising the steps of:

processing each of a plurality of unique bit patterns to generate signal sets;

transforming each of said signal sets to obtain a separate signal-defining pattern for each of said signal sets;

measuring a total error value between each of said separate signal-defining patterns and a preselected set of separate signal-defining patterns by performing a Euclidean distance measure operation for each of said separate signal-defining patterns;

distinguishing ones of said separate signal-defining patterns in which said measured total error values are greater than a predetermined error threshold;

adding said distinguished ones of said separate signal-defining patterns to said preselected set of separate signal-defining patterns, said preselected set of separate signal-defining patterns representing said plurality of distinct signal sets;

assigning one each of said plurality of distinct signal sets to one each of said first plurality of data subsets; and

associating each of said first plurality of data subsets with said assigned ones of said distinct signal sets in said first table.

27. A method as claimed in claim 26 wherein the digital data stream comprises video data.