Obviousness

Under 35 U.S.C. § 103, a patent claim is obvious if "the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains."

The present invention, US patent 8717203, relates to data compression systems and methods utilizing a combination of content-independent and content-dependent data compression. The core inventive concepts, as described in the "Summary of the Invention," include:

1. A method for compressing data by analyzing a data block to identify its data type; performing content-dependent compression if the data type is identified; and performing content-independent compression if the data type is not identified.
2. Specifics of content-independent compression, involving encoding with multiple encoders, determining compression ratios, comparing them to a threshold, and selecting the best-compressed block or the original block if no threshold is met.
3. Specifics of content-dependent compression, involving selecting one or more encoders associated with the identified data type, encoding, determining compression ratios, comparing them to a threshold, and selecting the best-compressed block or the original block if no threshold is met.
4. A method that performs content-independent compression if the data type is not identified or if the content-dependent compression does not meet a specified threshold.

Prior Art References:

The primary prior art reference explicitly discussed and contrasted within US8717203 itself is:

• Chu (U.S. Pat. No. 5,467,087), entitled “High Speed Lossless Data Compression System.” Chu discloses a data compression system where a data pre-compressor identifies the data type of an input stream and generates a data type identification signal. A data compressor then selects a data compression method from a preselected set of methods to compress the input data stream, "with the intention of producing the best available compression ratio for that particular data type". This system is described as a "content dependent high-speed lossless data compression and decompression system/method according to the prior art".

Limitations of Chu (as described in US8717203):

US8717203 explicitly identifies several limitations of the Chu method that it aims to overcome:

• The "need to unambiguously identify various data types," noting that data types may be "interspersed or partially compressed, making data type recognition difficult and/or impractical".
• The difficulty or impracticality, even "given a known data type, or mix of data types... to predict which data encoding technique yields the highest compression ratio".
• General problems with lossless compression, such as "content sensitive behavior," "significant variations in the compression ratio obtained when using a single lossless data compression technique," and "negative compression" where highly compressed data may expand.

Level of Ordinary Skill in the Art (PHOSITA):

A person having ordinary skill in the art (PHOSITA) in data compression at the time of the invention (priority date 1998-12-11) would be familiar with:

• Various lossless and lossy compression algorithms (e.g., run length, Huffman, Lempel-Ziv Dictionary Compression, arithmetic coding, data compaction, data null suppression, MPEG4, voice codecs, MPEG3, AC3, AAC).
• Concepts such as compression ratio, data expansion, and the impact of data content on compression effectiveness (data dependency).
• Basic system design principles for selecting and applying algorithms, including empirical comparison and the use of thresholds.
• The engineering practice of implementing fallback mechanisms when a primary method fails or produces undesirable results.

Obviousness Analysis under 35 U.S.C. § 103:

The core inventive contribution of US8717203 lies in its robust handling of situations where a purely content-dependent approach (like Chu's) is insufficient. The proposed combinations of prior art would render the claims obvious for the following reasons:

Combination 1: Chu (U.S. Pat. No. 5,467,087) + General knowledge of multi-algorithm, content-independent compression with empirical selection.

• Chu's Disclosure: Chu teaches analyzing data to identify its type and then selecting a compression method for that type to achieve the "best available compression ratio". This forms the basis for the content-dependent part of US8717203's claims.
• Motivation for Combination: US8717203 itself highlights that Chu's method struggles when data type recognition is "difficult and/or impractical". A PHOSITA, aiming to create a more robust and reliable compression system, would be highly motivated to implement a fallback mechanism for such scenarios. When a data type cannot be identified, the most straightforward approach to attempt compression is to use a "content-independent" strategy. This involves applying a plurality of known encoders (e.g., run length, Huffman, Lempel-Ziv, which are described in US8717203 as "well known within the art") to the data block, empirically determining the compression ratio for each, and then selecting the one that yields the best compression (or outputting the original data if none meet a minimal threshold). This multi-encoder, best-result-wins approach is a common, non-inventive engineering solution for achieving optimal compression when data characteristics are unknown or unpredictable.
• Reasonable Expectation of Success: Combining Chu's content-dependent approach with this general knowledge to address the stated limitation (unidentified data types) would have a high expectation of success, as it involves integrating well-known techniques.

Combination 2: Chu (U.S. Pat. 5,467,087) + General knowledge of fallback mechanisms based on performance thresholds.

• Chu's Disclosure: Chu aims for the "best available compression ratio" for a given data type.
• Motivation for Combination: Even if Chu's system correctly identifies a data type and selects an associated compression method, there is "significant variation in the compression ratio" and the risk of "negative compression" (data expansion). A PHOSITA would be motivated to ensure that the compression process is always beneficial. Therefore, it would be obvious to incorporate a performance check: after applying the content-dependent compression, determine its actual compression ratio and compare it against a predefined threshold. If the compression ratio is unsatisfactory (e.g., below a minimum threshold, indicating expansion or negligible gain), a PHOSITA would logically implement a fallback. The most obvious fallback would be to either output the original, uncompressed data (appending a "null compression descriptor") or to proceed with a content-independent multi-encoder approach (as described in Combination 1), which is designed to find the best empirical compression regardless of data type. This combination directly addresses the reliability and effectiveness shortcomings of a purely predictive content-dependent system.
• Reasonable Expectation of Success: The use of performance thresholds and conditional logic to manage data processing based on those thresholds is a fundamental concept in software and hardware engineering, making this a straightforward integration with high expectation of success.

Conclusion:

The methods and systems claimed in US8717203, particularly those related to the conditional application of content-independent compression when content identification fails or content-dependent compression performs poorly, would have been obvious to a PHOSITA at the time of the invention. The patent itself identifies the precise limitations of Chu's prior art that its invention purports to solve. The solutions—employing known multi-algorithm compression strategies, empirical performance evaluation, and threshold-based decision making—were conventional engineering approaches readily available to a PHOSITA to address such known problems, particularly when aiming to achieve robust and optimal compression across diverse data streams.