Apparatus, system, and method of data compression

• US 7,280,052 B2
• Filed: 09/30/2005
• Issued: 10/09/2007
• Est. Priority Date: 09/30/2005
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Levy, Amir
• Primary Examiner(s): Jeanglaude; Jean Bruner
• Application Number:
  US11/239,474
• Publication Number:
  US 20070085714A1
• Time in Patent Office: 739 Days
• Field of Search: 341/50, 341/51, 710/2, 710/202, 710/101, 360/69
• US Class Current: 341/51
• CPC Class Codes: H03M 7/30 Compression speech analysis-synthesis for redundancy reduction G10L19/00; for image communication H04N; Expansion; Suppression of unnecessary data, e.g. redundancy reduction

Abstract

A method and system of compressing an input data item such as, e.g., a cache line of machinc-executable code, by selecting a partition of the data item into subsequences that appear as words in a dictionary of a probability-based encoding scheme, such that the encoding of the subsequences produces a compressed representation of the data according to a desired efficiency parameter, e.g., minimized length. The method may implement an algorithm to identify one or more possible compressed representations of the input data item and to select a more efficient compressed representation.

Claims

1. A method comprising:

storing in a compression array a first possible representation of an input data item;

comparing a first value of a compression efficiency parameter relating to one or more subsequences of said first representation, with a second value of said compression efficiency parameter relating to one or more subsequences of a second possible representation of said input data item; and

replacing in said compression with said one or more subsequences of said first possible representation with said one or more subsequences of the second possible representation if said comparison indicates that said second possible representation is more efficient for compression of said input data item;

wherein the first and second possible representations correspond to first and second partitions, respectively, of said input data item into subsequences looked up in a compression dictionary.

2. The method of claim 1, comprising compressing said input data item based on the content of said compression array.

3. The method of claim 1, comprising looking up said one or more subsequences of said input data item in said compression dictionary to provide said partitions of said input data item.

4. The method of claim 3, comprising encoding said one Or more subsequences of said first and second partitions, using an encoding tree associated with said compression dictionary, to provide said first and second representations, respectively.

5. The method of claim 1, wherein comparing said first aid second values comprises comparing a value relating to the length of an encoding of said one or more subsequences of said first representation and a value relating to the length of an encoding of said one or more subsequences of said second representation, respectively.

6. The method of claim 5, wherein replacing in said compression array comprises replacing if said second compression efficiency parameter value corresponds to a shorter encoding length than said first compression efficiency parameter value.

7. An apparatus comprising:

a compression engine to store a first possible representation of an input data item corresponding to a first partition of said input data item into one or more subsequences of said input data item, to compare a first value of a compression efficiency parameter relating to said first representation with a second value of said compression efficiency parameter relating to one or more subsequences of a second possible representation of said input data item, and to replace said one or more subsequences of said first possible representation with said one or more subsequences of the second possible representation if said comparison indicates that said second possible representation is more efficient for compression of said input data item.

8. The apparatus of claim 7, wherein said compression engine comprises a compression array having a plurality of cells to store said compression efficiency parameter values relating to said first and second representations.

9. The apparatus of claim 8, wherein said compression engine is to compress said input data item based on the content of said compression array.

10. The apparatus of claim 7, wherein said compression engine comprises a compression dictionary having entries corresponding to a plurality of subsequences of said input data item having a high recurrence probability.

11. The apparatus of claim 7, wherein said compression engine comprises an encoding tree associated with said compression dictionary to provide an encoding of said dictionary entries such that a subsequence with a relatively high recurrence probability may be associated with an encoding of relatively short length.

12. The apparatus of claim 7, wherein said compression engine is to replace said one or more subsequences if said second compression efficiency parameter value corresponds to a shorter encoding length than said first compression efficiency parameter value.

13. A system comprising:

a memory having stored thereon at least a compression dictionary having entries corresponding to a plurality of subsequences of an input data item having a high recurrence probability; and

a compression engine to store a first possible representation of an input data item corresponding to a first partition of said input data item into one or more subsequences of said input data item looked up in said encoding dictionary, to compare a first value of a compression efficiency parameter relating to said first representation with a second value of said compression efficiency parameter relating to one or more subsequences of a second possible representation of said input data item, and to replace said one or more subsequences of said first possible representation with said one or more subsequences of the second possible representation if said comparison indicates that said second possible representation is more efficient for compression of said input data item.

14. The system of claim 13, wherein said memory is to further store an encoding tree associated with said compression dictionary to provide an encoding of said dictionary entries such that a subsequence with a relatively high recurrence probability may be associated with an encoding of relatively short length.

15. The system of claim 13, wherein said compression engine is to replace said one or more subsequences if said second compression efficiency parameter value corresponds to a shorter encoding length than said first compression efficiency parameter value.

16. The system of claim 13, wherein said compression engine is to produce a compressed version of said input data item based on the content of said compression array, to produce a compressed version of said input data item.

17. The system of claim 16, comprising a memory to store said compressed version of said input data item.

18. The system of claim 16, comprising a decompression engine including a decompression dictionary having entries corresponding to the entries of said compression dictionary, wherein said decompression engine is to decompress said compressed version using said decompression dictionary.

19. The system of claim 18, wherein said decompression engine comprises a decoding tree associated with said decompression dictionary, to provide a decoding of said compressed version into one or more subsequences corresponding to said decompression dictionary entries.

20. A machine-readable medium having stored thereon data compressed by:

storing in a compression array a first possible compressed representation of an input data item;

comparing a first value of a compression efficiency parameter relating to one or more subsequences of said first representation, with a second value of said compression efficiency parameter relating to one or more subsequences of a second possible compressed representation of said input data item; and

replacing in said compression array said one or more subsequences of said first possible representation with said one or more subsequences of the second possible representation if said comparison indicates that said second possible representation is more efficient for compression of said input data item,

wherein the first and second possible compressed representations correspond to first and second partitions, respectively, of said input data item into subsequences looked up in a compression dictionary.

21. The machine-readable medium of claim 20 having further stored thereon a decompression dictionary having entries corresponding to the entries of said compression dictionary, said decompression dictionary adapted to be used by a decompression engine to decompress said compressed data.

22. The machine-readable medium of claim 21 having further stored thereon a decoding tree associated with said decompression dictionary, to provide a decoding of said compressed representation into one or mote subsequences corresponding to said decompression dictionary entries.