What is Data Compression?
Data compression is the process of removing redundancies from data in such a way that the key information contained in the data is preserved. In lossless compression every bit of the original data can be recovered perfectly after decoding. Otherwise, the compression is described as being "lossy".

Why do I need Data Compression?
The immediate benefit of data compression is the reduction in size of the data, which translates to savings in storage space and transmission time. The space and time reduction usually lead to lower operating costs (both at the personal and enterprise levels) as storage hardware and bandwidth are put into more optimal usage.

Is Miliki the same technology as JPEG2000?
No. Miliki is designed to address high compression of a variety of data types including documents (e.g., Excel, PowerPoint, Word, PDF, etc) and pictures (e.g. JPEG, BMP, GIFS, Animated GIFS, PNG, TIFF, etc). Miliki does not use just one compression algorithm. Miliki employs a compression approach that utilizes a suite of algorithms and filters that are optimized for particular data types. Over 30 different compression filters and algorithms can be called to optimally compress a specific data type.

How is the high compression result achieved in Miliki?
Miliki uses a Content Sensitive Approach to identify and decompose a data type to suitable constituents for effective compression. For example, an image is automatically grouped into one of many classes of pictures such as photo realistic, natural, line drawing, bi-tonal etc. Documents are also decomposed into constituent classes such as text, tables, formulas, embedded pictures etc. Miliki then selects from the suite of algorithms and filters the best scheme for compressing each component. The result is the significant compression that is much better than what is produced by popular compression techniques.

Has there been any independent laboratory evaluation of the compression engine of Miliki?
Yes. National Software Testing Labs (NSTL) conducted a test of the LAFE Technologies Optimizer which uses the Miliki compression engine. The PDF version of the report can be obtained at http://www.LAFEnet.com/download/nstl_report.pdf.

Are the compression filters and algorithms proprietary?
A large class of the compression filters and algorithms are proprietary to LAFE Technologies. The company holds a number of patents on data (including images, audio and video) compression, encryption and accelerated data delivery across communications networks. Some of the image compression schemes use sophisticated models that are optimized for the human vision system. Since Miliki must also be able to read and write all the data types that are supported, the product evidently contains non-proprietary libraries that are utilized for some of the read/write purposes.

What other data types will Miliki support in future versions?
Future versions will incorporate compression of multimedia files including AVI, MPEG, WAV and MP3 files. A special Miliki edition designed specifically for further compression of ZIP data is also planned for the future.

Why is the Miliki compression time longer than that for ZIP for some files?
Miliki uses over 30 different compression schemes (versus just 1 for ZIP). The application has to analyze every data file and decide on the best scheme for different parts of the data. This is where there is a small penalty for the encoding time. For example, in the case of an Excel or a PDF file, Miliki decomposes the file into its component images, text, formulas, tables etc. The optimal compression scheme is then selected for each of these components. There is a cost paid in processing time for this decomposition process. The ultimate benefit is the reduction in the size of the compressed data. Furthermore, the extra processing time is a tiny fraction of the time it would normally take to transmit a much larger file through a narrow bandwidth. The supreme compression achieved with Miliki saves transmission time and optimizes the space on storage devices.

For certain files, why does the recovered data have a size different from the original?
Miliki always ensures that the integrity of the recovered data matches that of the original. The discrepancy in the recovered size occurs when the component parts (which were compressed using one of the over 30 special schemes) are reassembled and decompressed into the original format. This is especially true when the original data includes lossy formats such as JPEG images. Furthermore, Miliki will remove some garbage that might have been created by the original application (eg Powerpoint) that created the data. This garbage has nothing to do with the information content of the data. But in all cases the integrity of the original data (e.g., 1000 will not become 2000) is preserved.

What are Cellular Automata Transforms?
Cellular Automata Transforms (CAT), invented at LAFE Technologies, are computational operations applied to the evolving field of cellular automata. Cellular Automata (CA) are dynamical systems in which space and time are discrete. The cells are arranged in the form of a regular lattice structure and each must have a finite number of states. These states are updated synchronously according to a specified local rule of interaction. Cellular Automata has been a subject of interest in several circles. Stephen Wolfram’s book – A New Kind of Science (2002), contains description of the essence of CA and how they can serve as suitable models of natural processes.

How are CAT used for processing data?
At LAFET Technologies, we have discovered a smart approach to addressing the astronomical number of rules describing a large family of cellular automata that can be used for data encoding applications including data compression, data encryption and the modeling of complex processes. CAT involves the generation of information building blocks. Each building block has an associated cellular automata rule. A given data can be represented by a collection of these information building blocks. The size of the CA rules for the building blocks is smaller than that of the original data. The use of the information building blocks therefore results in a more compact representation (i.e., compression) of the data. Only the associated rules used for generating the building blocks need be stored or transmitted in the place of the original data. Furthermore, because there are more CAT rules than the age of the universe, the coding of any data using these rules is also an effective way of encrypting the data. A code breaker will have to employ the services of the most powerful computers, working longer than the known age of the universe to decipher a message that has been encrypted with CAT information building blocks.

How can I find out more about CAT?
For a more detailed information on CAT please refer to Dr. Lafe's book: Cellular Automata Transforms -Theory and Applications in Multimedia Compression, Encryption, and Modeling circulated by Kluwer Academic Publishers, Boston/Dordrecht/London, 2000. The book is available at www.amazon.com and many e-commerce stores.