MINC LOSSLESS DIGITAL DATA COMPRESSION SYSTEM

An Overview of the Technical and Marketing Aspects

Of MINC Data Compression System



By Jean Simmons


President and CEO of Premier America Corporation


September, 1997

 

(The contents of this document are Proprietary and Copyrighted Material of Premier America Corporation)

 

 

TABLE OF CONTENTS

  

INTRODUCTION

DEFINITION OF MINC

    • No Limit Concept
    • An Iterative Algorithm
    • Compressing Random Data
    • MINC is Lossless

MINC FEATURES

    • "No Limit"
    • Lossless
    • Universality
    • Compresses the Compressed

MINC ADVANTAGES

    • Unlimited Bandwidth
    • Unlimited Storage Capacity
    • Unlimited Rates of Data Transmission

 MINC BENEFITS

    • Technical
    • Economic

A BRIEF HISTORY OF DATA COMPRESSION

THE BUSINESS BEHIND MINC

CONCLUSION

ABOUT THE AUTHOR 

DISCLAIMER

 APPENDIX

 

 

MINC LOSSLESS DIGITAL DATA COMPRESSION SYSTEM

 

 An Overview of the Technical and Marketing Aspects

Of MINC Data Compression System

 By Jean Simmons

 September, 1997

 

(The contents of this document are Proprietary and Copyrighted Material of Premier America Corporation)

  

INTRODUCTION

  • In 1996 over 200,000 people converged upon a desert oasis. The visitors had cellular phones; as did the residents. At some point during the worlds largest computer convention, held in Las Vegas, a technological nightmare occurred. Despite the best efforts of cellular service providers, the wireless network could not cope. The most modern and sophisticated equipment in the world failed to prevent the airwaves from clogging. Many people could not get through at all. Those who did, found their cellular phone transmissions were lost, dropped, or interrupted.

 

  • For most attendees the Comdex incident of 1996 highlighted the urgent need for a breakthrough to provide more bandwidth for cellular phones and other wireless products. Consumers ever growing desire for faster access to better communications and more information continues to place enormous demands on technology manufacturers to keep pace.
  • The solution to this urgent need comes in the form of a newly developed "no-limit" lossless digital data compression method that its developer, Premier Research, LLC, calls "a quantum leap in compression technology".
  • Its name is MINC Lossless Digital Data Compression System. MINC is an acronym for Minimizing INformation Coder.

  

Before proceeding, please note that a compilation of relevant terminology has been provided as an aid in understanding the concepts behind MINC compression system. See Appendix I.

 

Definition of MINC

 

  • MINC is a lossless digital data compression system that uses an iterative process to achieve "no limit" data compression.

 

"No limit" lossless digital data compression means:

    • "No limit" on the amount of compression which can be achieved
    • "No limit" on the type of information which can be compressed

 

An Iterative Algorithm

  • The MINC algorithm is designed to be used iteratively. That is, the input data is compressed slightly in a single step which is called a "cycle". The compressed output produced by this first step or first cycle is then compressed again in a second cycle using exactly the same procedure followed in the first. The effect of this iterative process is to progressively build an ever-increasing ratio of compression relative to the original input, one small step at a time. On average, the MINC algorithm achieves about 25% compression at each cycle (iteration). Therefore, 8 cycles yield 10:1 compression. But each 8 cycles thereafter yields another power of 10 as shown below:

Number of Cycles

8

16

24

32

40

48

etc.

Compression Ratio

10:1

100:1

1,000:1

10,000:1

100,000:1

1,000,000:1

etc.

  • If the application requires another order of magnitude of compression, just add another 8 cycles. It is because of this we say that there is "no limit" on the amount of compression, which MINC achieves.

 

  • If at any cycle the size of the compressed data becomes too small to continue through further cycles, it can simply be batched with other compressed data and the process continued. The practical constraint on how far this process continues becomes one of time and is, therefore, application dependent.

 

  • When the MINC algorithm is implemented in a silicon chip, using a parallel architecture known as a "systolic pipe", the time issue will be a moot point because most of the processing time will be overlapped.

 

Compressing Randomized Data

  • It will be apparent from the foregoing discussion that as data proceeds downward through the various cycles (iterations), the data is becoming increasingly randomized. That is, the entropy of the data is increasing. Any patterns present in the source data are rapidly destroyed.

 

  • The MINC algorithm is unaffected by this increasing randomness. Most lossless compression algorithms depend upon the fact that the source symbols occur with a definite frequency distribution that permits shorter codes to be assigned to more frequently occurring symbols, and longer codes to less frequently occurring symbols. The net result is compression. Hence, such methods cannot compress random data. In contrast, the MINC algorithm is not concerned with the frequency of occurrence of source symbols. In fact, MINC requires no definition of the source data whatsoever and, therefore, has no knowledge of the meaning attached to the source symbols. The MINC algorithm is designed to uncover order within chaos and structure underlying randomness. Because of this, MINC is unaffected by the increasing randomness of the data as we proceed downward in the cycles. Therefore, the number of iterations can continue virtually indefinitely. This is the basis for MINC s claim of "no limit" on the amount of compression.

 

MINC is Lossless

  • The data compression world is divided into 2 camps: lossy and lossless

 

Lossy techniques are constructed so as to allow decompressed data to differ from the original data so long as the decompressed data satisfies some fidelity criteria. For example, with image or audio compression (such as found in MPEG and JPEG) it may suffice to have an image that looks as good to the human eye or sounds as good to the human ear as the original. In other words, much of the original information can be discarded before the eye or ear detects it. As a result, lossy compression methods, although they can achieve large ratios, can be used only with certain data types such as video and audio.

 

In contrast, with lossless techniques the decompressed data is identical to the original data bit-for-bit. Nothing is altered and the source data is restored with 100% integrity. Most kinds of data compression (excluding video and audio information) require lossless techniques. Historically, lossless methods have only achieved an average of about 2:1 compression. This is not enough to meet the growing demands for more storage capacity, bandwidth and faster transmission rates.

 

  • MINC solves all this. It is lossless and achieves very large compression ratios as we have discussed previously. This means that MINC can be used with all types of information.

 

 

MINC FEATURES

  

  • Here are the main features of MINC Lossless Digital Data Compression System which taken together set MINC apart from all current and prior art in this field.

 

    • There is no limit on the amount of compression MINC can achieve.
    • MINC is a lossless digital compression method.
    • MINC is universally applicable to all data types.
    • MINC can compress data already compressed by other methods.

 

No Limit

  • The iterative procedures utilized by MINC mean that there is virtually no limit on the amount of compression that can be achieved. If additional compression is required, then more iterations are performed. The increasing randomization of the data which results from these iterations in no way diminishes MINC s ability to achieve further compression.

 Lossless

  • Because MINC employs "lossless" rather than "lossy" techniques, it is able to recover all compressed data bit-by-bit with no loss of the original information. Decompressed data exactly equals the original.

 

Universality

  • The methods employed by MINC to achieve compression are applicable to all digital information. There is no need to tailor or adapt the MINC algorithm for different data types. There is also no need for prior definition of data types. This facilitates the use of MINC in a real-time communications environment.

 

Compresses the Compressed

  • MINC is able to compress data already compressed by other methods. This means MINC is compatible with existing compression methods and standards. Present methods and standards cannot make this claim.

  

MINC ADVANTAGES

  

  • The MINC features described in the previous section provide the following advantages for applications and products in which MINC shall be embodied.

 

    • Unlimited Bandwidth
    • Unlimited Storage Capacity
    • Unlimited Rates of Data Transmission

 

Unlimited Bandwidth

  • When MINC is used in a telecommunications environment, its very large compression ratios effectively provide unlimited bandwidth. Whatever bandwidth has been assigned to an application, with MINC that same bandwidth will allow a vast increase in the amount of information transmitted. The effect is the same as if there had been a substantial increase in bandwidth. Thus, MINC ends forever bandwidth and frequency congestion issues.

 Unlimited Storage Capacity

  • The large compression ratios provided by MINC effectively multiply the capacity of any storage system with which it is used. Since the space required to store any item of data compressed with MINC is a small fraction of what it would otherwise occupy, the effect is equivalent to increasing the capacity of the storage device. This applies to all types of storage devices and media (e.g. hard disk, floppy disk, magnetic tape, CD-ROM, CD, videotape, etc.).

 Unlimited Rates of Data Transmission

  • Since data is compressed prior to transmission, the amount of time to send compressed data over a communication channel (e.g. phone line, cellular channel, etc.) is reduced by the same proportion as the compression ratio. For example, if the compression ratio is 100:1, then the transmission time is 1/100th of what it would be without MINC . Since MINC provides practically unlimited compression ratios as discussed above, this translates into virtually unlimited transmission rates.

 

MINC BENEFITS

  • The MINC features and advantages discussed in the preceding sections provide many benefits to equipment manufacturers, service providers and end users. We have grouped the benefits into two categories:
    • Technical benefits
    • Economic benefits

 

Technical Benefits

  • MINC is compatible with all existing digital technology for both data storage and telecommunications. Yet MINC effectively removes all of the inherent limitations of these technologies today: limitations of (1) storage capacity (2) bandwidth (3) data transmission rates. Therefore, MINC permits a giant "leapfrog" in technical benefits and cost-performance improvements to existing products. The demand for increased data storage capacity accelerates constantly, particularly with multimedia applications. MINC solves this problem since it has the ability to increase compression ratios as needed to meet future demands. The same applies to the ever-increasing demand for bandwidth. MINC satisfies this demand by the ability to increase compression ratios to allow available bandwidth to carry the required volume of traffic. It does this without the need for laying another foot of copper wire, coaxial cable, or fiber optics. In fact, MINC raises the question of whether in the future we will need wire-based communication at all. MINC endows available frequency spectrum with the capacity to carry all present and future telecommunication traffic in wireless form. This includes TV and radio broadcasting, mobile radio, cellular, paging, fax, voice and data.

 

Economic Benefits

  • All the technical benefits outlined above are available to manufacturers and service providers without any extra allocation of resources for research and development. This will cost-justify the modest royalty for the use of MINC , in exchange for which, they receive a complete technology ready to be interfaced with their systems. Since MINC expects to receive a digital bit stream as its input and produces the same as its output, interfacing existing systems to MINC is trivial. To facilitate integration with existing products, a family of semiconductor chips will be available from licensed MINC manufacturers for incorporating MINC into current and future products. Preliminary technical specifications will be available to licensees shortly. Obviously, the benefits of unlimited bandwidth, unlimited storage capacity and accelerated data transmission rates will bring huge cost performance gains to all systems into which MINC is incorporated. The fact is that MINC will facilitate all telecommunications being in wireless form and will save trillions of dollars in capital investment in expanding and converting wire networks (e.g. installing fiber optics).

 

A Brief History of Data Compression

 

    • In 1952, David Huffman of Bell Labs published a method for lossless data compression. His method guaranteed optimum codes for any data set with a known frequency distribution. Those symbols in the data set (the source alphabet) which occurred most often were assigned the shortest codes. Those symbols, which occurred less often, were assigned longer codes. This method on average achieved 2:1 compression. In the last 45 years, most lossless methods have been based on Huffmans work. Even today, most lossless compression methods achieve an average of only 2:1.
    • In recent years, much work has been done in the area of lossy compression for video and audio information. This has given rise, for example, to the MPEG standard for video. Lossy compression based on the use of fractals has also appeared. These methods are inherently limited by the fact that they degrade the quality of the restored data.
    • MINC represents a new chapter in the history of data compression. It is lossless and yet provides no limit on the compression which can be achieved. MINC is universally applicable to all digital information. It needs no prior definition of the data to be compressed and is therefore ideal for real-time environments. There is no longer any need for using lossy methods.

  

 

THE BUSINESS BEHIND MINC

    • Premier Research, LLC, a privately held Nevada company based in Las Vegas, is the owner and developer of the technology known as MINC . Patents pending covering MINC have been assigned to Premier Research, LLC.

 

    • Premier America Corporation, a Nevada corporation, also based in Las Vegas, is the exclusive worldwide marketing agent of MINC . An agency agreement exists between Premier America Corporation and Premier Research, LLC. Jean Simmons, the author of this paper, is the President and CEO of Premier America Corporation.

 

    • The MINC technology will be made available on the basis of non-exclusive technology licenses. Licensees will pay a modest royalty for the use of the technology to Premier Research, LLC. These licenses will be marketed and the terms negotiated on behalf of Premier Research, LLC by Premier America Corporation.

 

    • It is intended the MINC technology will be implemented as a family of semiconductor chips to be manufactured by MINC licensees. These chips will be available for purchase and incorporation into products of suitably licensed manufacturers.

 

CONCLUSION

    • To meet the demands of the dramatic growth in the quantity of digital information that must be transmitted, stored, and managed in the information age, new technology is required.

 

MINC , by providing:

    • Unlimited bandwidth;
    • Unlimited storage capacity;
    • Unlimited rates of data compression;

is that technology!

 

 

It is here NOW.

It WORKS.

  

For more information contact:

Jean Simmons

President and CEO

Premier America Corporation

2860 East Flamingo Rd. Suite A

Las Vegas, NV 89121

702-737-6462

Fax 702-737-1789

 

Send E-mail

 

ABOUT THE AUTHOR 

  • JEAN SIMMONS, President and CEO of Premier America Corporation, has nineteen years of management-level experience in sales, marketing, and operations while working for various high-technology companies. Her career as a management-level executive included positions at NYNEX Corporation, Toshiba America, TRW, Inc. and Xerox Corporation. Currently, Simmons lives in Las Vegas and, as President of Premier America Corporation, is actively initiating a national marketing/sales program for MINC Data Compression System.

 

DISCLAIMER

 

  • Premier America Corporation based on currently available information believed to be reliable has prepared this document. Premier America Corporation, nor any of its respective directors, officers, employees or representatives, make any representation or warranty, expressed or implied, as to the accuracy or completeness of this document or any of its contents, nor shall any of the foregoing have any liability resulting from the use of the information contained herein or otherwise supplied.

 

 

APPENDIX I - TERMINOLOGY

 

  • ALGORITHM a procedure for solving a mathematical problem, in a finite number of steps that frequently involves repetition of an operation.

 

  • BANDWIDTH refers to a range of frequencies which have been assigned a particular use for communication purposes.

 

  • BINARY NUMBERS numbers expressed in (base-2) notation, a system that uses only two digits, 0 and 1.

 

  • CYCLE as used here, a single iteration in the process of compressing or decompressing information.

 

  • DATA COMPRESSION refers to the process of transforming a body of data to a smaller representation from which the original or some approximation to the original can be computed at a later time.

 

  • DECODING the act of converting coded information into its original form.

 

  • DIGITAL DATA representing information (data) as electrical "on-off" signals that correspond to binary digits and can be stored in computer memory.

 

  • ENCODING the act of converting information into a code.

 

  • ENTROPY is a measure of the information content in a message.

 

  • ITERATION repetition of a sequence of computer instructions a specified number of times or until a condition is met.

 

  • LOSSLESS DATA COMPRESSION requires all data compressed and subsequently decompressed to always be identical to the original.

 

  • LOSSY DATA COMPRESSION allows the decompressed data to differ from the original data so long as the decompressed data satisfies some fidelity criteria.

 

  • RANDOM NUMBERS numbers distributed as though by chance, usually with equal probability of occurrence.

 

  • REDUNDANCY the part of a message that can be eliminated without loss of essential information.

 

  • STORAGE DEVICE a memory device in which data is stored for use by a computer system.

 

 

APPENDIX II - DEBUNKING THE MYTHS

 

 

  • MINC appears to contradict the conventional wisdom of data compression. There are good reasons for this. Traditional methods of lossless data compression, practiced for over 45 years, have given rise to this conventional wisdom. It must be understood that MINC departs radically from these methods and, therefore, runs counter to widely accepted notions.

 

  • In this section, we seek to debunk the old myths and show that MINC requires new ways of thinking about old problems.

 

MYTH # 1 - A Universal Lossless Data Compressor

(i.e. a compressor of everything) is mathematically impossible.

 

  • It is maintained by some that a universal lossless data compression algorithm is not possible on mathematical grounds. By "universal" we mean an algorithm, which can compress anything including random numbers. These critics also state that if such claims are made for any algorithm, it is not necessary to know the details of the algorithm in order to reject it on mathematical grounds.

 

  • The simplest basis for mathematical objections to universal lossless data compression is the "counting" argument. The essence of this argument is that one cannot represent Sn things uniquely with n-l bits. Of course, as it stands, this statement is true and any assertion to the contrary is nonsensical. However, this statement overlooks a very simple and very important mathematical fact:

 

Namely that Sn = Sn-1 + Sn-2 +. . .+Sn-n +1.

We give an example of the above equation as proof.

Let S = 2 and n = 8.

28 = 256

 

Summing all powers of 2 less than 8 + 1 gives:

 

2 7 = 128

2 6 = 64

2 5 = 32

2 4 = 16

2 3 = 8

2 2 = 4

2 1 = 2

2 0 = 1

+ 1

256

 

  • The importance of the above is that it shows that Sn things can be uniquely represented by a set of variable length binary strings each of which is shorter than n bits by at least 1 bit.

 

  • The MINC universal lossless data compression system is based on this simple foundation. Of course, this leaves open the question of how to accommodate the representation of the length of these strings and still realize net compression.

 

  • The MINC algorithm has solved this problem. Once again this contradicts the pundits, since to evaluate this issue one must know just how MINC resolves it.

 

MYTH # 2 - There is not sufficient redundancy in most real-world data to get the large compression ratios MINC provides.

 

  • When considering the issue of redundancy in a data set versus its true information content, it is customary to overlook the staggering amount of redundancy introduced by the process of digitizing information.

 

  • Consider the example of English text. In digital form each letter in a block of text occupies 1 byte (8 bits). An average length word (e.g. 5 letters), therefore, occupies 5 bytes (40 bits). A code of this length can represent 240 things (1,000,000,000,000). However, Oxfords Unabridged Dictionary contains only about 219 words (500,000). Typically, an average word of English text when digitized is about 2 million:1 redundant. The absurdity of this becomes even more apparent if we consider sentences. A 7-word sentence could occupy perhaps 280 bits (7 words at 40 bits per word). A code of this size (2280 or 1,000,000,000,000,000,000,000,000,000,000, 000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000) can represent a number of things greater than there are protons in the universe.

 

  • There are other examples of such digital absurdities. Consider digital speech. The digitization of speech with pulse code modulation (PCM) and 8-bit samples with 8,000 samples per second produces 64,000 bits to represent 1 second of human speech. This means we are assigning a code capable of representing 264,000 things to the information content in 1 second of human speech.

 

  • Formal methods do not exist for removing this kind of redundancy. However, the MINC algorithm does eliminate this redundancy using its iterative procedures. It should not be surprising, therefore, that MINC achieves the very large compression ratios that it does.

 

MYTH # 3 - Randomized data cannot be compressed.

 

  • Traditionally, lossless data compression methods have relied on the fact that for most real-world data sources, the frequency of occurrence of the source symbols varies widely. Such methods exploit this fact by assigning shorter codes to the most frequently occurring symbols and longer codes to those that occur less often. However, randomized data symbols tend to occur with approximately equal frequency. Therefore, traditional methods are unable to compress such information.

 

  • By contrast, MINC is not concerned with the probable frequency of occurrence of the source symbols. MINC does not examine the source data in fixed-length increments (e.g. bytes). Such bit groups for random data will, of course, occur with approximately equal frequency. Instead, MINC treats the source data as composed of variable-length bit sequences, which by definition have certain properties that will always occur with absolute certainty. Predictability means compressibility. In this manner, MINC uncovers order within chaos and structure underlying randomness. For proprietary reasons, the nature of these sequences and their properties cannot be disclosed here.

 

MYTH # 4 - Data that has already been compressed cannot be compressed further.

 

  • The reason for this is closely related to MYTH # 3. Compressed data has been at least partially randomized and cannot be further compressed by conventional methods. Once again, the randomizing of the data has no effect on MINC so it can continue to compress data which has been previously compressed by other methods. Because of this, MINC is compatible with existing compression methods and standards.

Bron: http://www.pacminc.com/white_paper.html (html closed)