FGD Data Compression Algorithm

   FGD is a fresh algorithmic approach for block-oriented lossless data compression. The algorithm takes into account the limitations and characteristics of small systems, such as low-memory requirements, high speed and short latency, as required by communication applications. FGD can be described by the categories of Lossless and Adaptive. Lossless since, while decompressing, it returns the data identical to the original source. Adaptive means that, for every data stream, it constructs an "ad-hoc" dictionary that reflects the particular characteristics of the data type.

   Block-compression is also different to the type of file-compression used in, for example, WinZIP. Those archivers compress the whole file as one entity (and probably several files at once, also), and thus require all the file history to decompress the last byte; no matter how long the file could be. On-line compression tasks usually can not afford such long delays, and commonly involve a more fixed block of input data do deal with.

   Compression rates, of course, vary depending on the compressibility of the data itself; but as a rule, an average coefficient of 2.5x (compared to 2.0x currently offered) is reached. As it is known, for different types of files, ZL1 and ZL2 usually alternate as the relative winner. RFGD strongest point is reaching better compression in all categories, the greater differences obtained for the more compressible data types. This statement can be confirmed by several Image and Data tests.

   On the basis of this technology GEMAC had developed the following Hardware Compression Solutions:

   For an algorithm to be on-line, it has to be One-Pass and fulfil the speed requirements of the system. One -pass are called those algorithms that process the data just once, as they entered the input stream, and their compression result is obtained based only on the previous existing cumulative information. On the contrary, those algorithms that gather first statistics on the data stream and later optimize the encoding based on a second (or n)-pass over the data are commonly known as Multiple-pass, and are the ones normally employed by current software archiver utilities like WinZip.

Advantages of FGD Data Compression Technology

• Hardware portability to different platforms
• Standard RAM for dictionary construction 
• Heterogeneous systems, consisting in both data and images could benefit from the unified approach, without the need to add different compression hardware modules for each data type, simplifying the overall design and cost.

   There are broad areas of application of this technology. We mention below only the most visible ones.

• Wireless and Wide-area communication devices. The doubled/tripled improvement in channel capacity represents a paramount characteristic, specially for those telecom devices that should operate under physical and infrastructure limitations.

• All types of storage media, but preferentially those that receive the advantages of improved performance, due to superior transfer rate, as is the case with tape storage and mobile storage media; as a premium factor additional to the default volume capacity increase.

• As a general lossless encoder unit, GEMAC technology can be tuned to specific multimedia purposes. As an example, with the help of image pre-filtering, GEMAC compression technology have been adapted to gray-scale and full-color images; obtaining typically compression rates of 3:1 without losses, satisfying the demands of the medical industry and other applications where legal or quality requirements demand the use of a fully reversible  compression.

• There is a natural marriage of GEMAC high-speed data compression solutions with encryption devices, offering overall performance improvements and higher security.

Software emulation

• ANSI-C compression/decompression.

• Assembler-optimized for x86 and 386+ processors, (used for years in a disk-compression product) is available on request.