Obviousness — US Patent 9054728

To analyze the obviousness of US patent 9054728, we need to consider the prior art references and determine if a person having ordinary skill in the art (POSA) would have been motivated to combine them to arrive at the claimed invention. The patent itself lists US 6195024 B1 as a prior art keyword and a priority claim. Therefore, this patent will be a primary focus for our obviousness analysis.

Prior Art References:

From the patent information, US 6195024 B1 is listed as a priority claim from 1998-12-11. We will assume this patent is relevant prior art for the obviousness analysis.

The patent US9054728B2 describes several well-known lossless encoding techniques in its detailed description, including:

Run length encoding [cite: US9054728B2]
Huffman coding [cite: US9054728B2]
Lempel-Ziv Dictionary Compression [cite: US9054728B2]
Arithmetic coding [cite: US9054728B2]
Data compaction [cite: US9054728B2]
Data null suppression [cite: US9054728B2]

These techniques are explicitly mentioned as "currently well known within the art" in the description of the encoder module 30. [cite: US9054728B2]

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

A 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." (MPEP § 2141).

The independent claims of US9054728B2 focus on systems and methods for data compression that involve:

Receiving a data block and determining its size.
Encoding the data block with a plurality of encoders.
Determining compression ratios for each encoded block.
Comparing these ratios against a threshold.
Selecting and outputting either the original block (with a null descriptor) or the best-compressed block (with a compression type descriptor).
(For some claims) identifying the data type to select appropriate encoders (content-dependent compression).

Combination of US 6195024 B1 with General Knowledge in Data Compression

Given that the patent US9054728B2 explicitly states that several common compression techniques like Huffman, Lempel-Ziv, and run-length encoding are "well known within the art," [cite: US9054728B2] a POSA would be familiar with their individual operation and benefits.

Lempel-Ziv (LZ) compression: LZW (Lempel-Ziv-Welch) is a universal lossless data compression algorithm published in 1984, improving on the LZ78 algorithm. It became widely used in Unix systems and for image formats like GIF and TIFF. LZ algorithms work by finding repetitive character strings in an input and replacing them with encoded versions.
Huffman coding: Huffman coding is an entropy encoding algorithm for lossless data compression, developed in 1952. It assigns variable-length codes based on the frequency of characters, with more frequent symbols getting shorter codes.
Run-length encoding (RLE): RLE is a lossless data compression technique where runs of identical data values are stored as a single occurrence of the value and a count of its repetitions. It is particularly efficient for data with many repetitive sequences, like simple graphic images. RLE was patented by Hitachi in 1983.

The core idea of applying multiple compression algorithms, comparing their results, and selecting the best one, or applying no compression if none are effective, is a known approach to optimize compression. The motivation for a POSA to combine these known techniques would be to achieve better compression ratios across diverse data types, as different algorithms perform better on different kinds of data. For instance, RLE is good for repetitive sequences, while Huffman coding is effective for data with varying character frequencies.

The notion of "content-dependent" compression, where the data type is identified to select suitable encoders, would also be obvious to a POSA. For example, video codecs like MPEG4 or voice codecs are specifically designed for those data types. [cite: US9054728B2] A POSA would understand that certain compression algorithms are more suitable for specific data types (e.g., image compression algorithms for image data, text compression for text data) and would be motivated to identify the data type to apply the most effective algorithm.

Specific Obviousness Combinations:

1. Claims 1 (Method for Content Independent Data Compression) and 11 (Content Independent Data Compression System):

These claims involve encoding a data block with a plurality of encoders, determining compression ratios, comparing them to a threshold, and selecting the best-compressed block or the original block with a null descriptor.

Motivation to Combine US 6195024 B1 with general data compression knowledge: US 6195024 B1 (if it details various compression techniques or a system for applying them) combined with the general knowledge that different compression algorithms (like those listed in US9054728B2 as "well known," e.g., Huffman, Lempel-Ziv, run-length encoding) have varying effectiveness depending on the data, would motivate a POSA to employ multiple encoders. A POSA would readily understand that by running multiple known encoders on a data block and comparing their output, they could select the most efficient one for that particular block. The concept of a compression ratio is fundamental to data compression, and setting a threshold to avoid negative compression or insignificant gains is also a standard engineering consideration. Appending a descriptor is a logical way to indicate which (if any) compression was applied, enabling proper decompression.

2. Claims 7 (General Method for Data Compression with Data Type Identification) and 17 (General Data Compression System with Data Type Identification):

These claims build upon the previous ones by adding the step of identifying the data type of the input data block before encoding.

Motivation to Combine US 6195024 B1 and General Data Compression Knowledge with Data Type Recognition: A POSA would be well aware that the effectiveness of compression algorithms is often highly dependent on the type of data being compressed. For instance, a Lempel-Ziv type algorithm excels at text with repeating patterns, while a run-length encoder is efficient for simple graphics with long sequences of identical pixels. The patent itself mentions content-dependent encoders like MPEG4 for video or various voice codecs. [cite: US9054728B2] Therefore, recognizing the data type (e.g., text, image, audio, video) and then selectively applying a subset of known encoders or even a single, highly optimized encoder for that specific data type would be an obvious optimization for a POSA seeking to improve compression efficiency and/or speed. A "data file recognition list(s) or algorithm(s)" [cite: US9054728B2] is a straightforward implementation for such a task.

Conclusion:

The independent claims of US9054728B2, which describe applying multiple compression encoders, comparing their efficiency, and selecting the best result (or no compression), and particularly those claims that add data type identification to guide encoder selection, would likely be considered obvious to a person having ordinary skill in the art by combining prior art such as US 6195024 B1 with the widely known and documented principles of data compression, including specific algorithms like Huffman, Lempel-Ziv, and run-length encoding. The motivation would be to optimize compression performance (ratio, speed, or both) for diverse data types, a common goal in the field of data processing.