WO1986003037A1

WO1986003037A1 - Data compressing computer system and process thereto

Info

Publication number: WO1986003037A1
Application number: PCT/US1985/002252
Authority: WO
Inventors: Konstantine Beridze
Original assignee: Konstantine Beridze
Priority date: 1984-11-16
Filing date: 1985-11-15
Publication date: 1986-05-22
Also published as: JPS62501115A; EP0204763A4; EP0204763A1

Abstract

The prior art compression of a data in computer memory and transmission of compressed data has proved to be uneconomical and too slow. The present process uses dimensional co-ordinates of the place of the data as its value for coding purposes. The coding process, as shown by the flow chart of Fig. 2C, is based on a compression of the distances between "on" and "off" states of bits of data bytes as they progress through stages (34, 35, 36, 37 and 38) and a decoding process of Fig. 2C, as shown by other stages (40, 41, 42 and 43). The coding and decoding processes are controlled by a repeating control stage (39) that increases the compression of the data as many times as ordered. By using the repeated coding of the dimensioned co-ordinates of the place of the data in the transmitted/received data in the computer system or in the transmission system, the volume of the computer memory is increased and the communications are done more economically.

Description

DATA COMPRESSING COMPUTER SYSTEM AND PROCESS THERETO

TECHNICAL FIELD

The present invention relates to new and useful process and devices for coding and compression any numerical, alphabetical and special characters, or just numbers into memory systems and for transmission and reception the data in communication systems more economical way that used at now, and it relates particularly to repeated coding process and devices for compression the data into memory systems, communication systems and related devices that increase the level of compression respectively.

BACKGROUND ART Since the first computers were built, it became a trivial fact that most advantagious way of representation of the data in a computer memory was tho binary system of numbers, although the base 3 could give some enlargment in a computer memory, as well as an additional difficulty of electronic realization. It were also known some packed formats for a numerical data and some methods of compressing the data, using better density of coding, that gave some not significant improvement of the volume of tho data.

Such methods of compressing the data may be presented by following examples: 1. U.S. patent 3,717,851, John Cocke and others, Feb. 20, 1973 used groups of alphabetical characters (a prefix of a word) rather than as single characters (a letter) in order to compact the data.

2. U.S. patent 3,936,808, Joseph L. O'Neill, Jr., Feb. 3, 1976 used to compact the data for use by a video display monitor. Each coded repeat sequence specifies a repeat operation and a particular number of times that a display data character is to be repeated in a display line. The invention compacts the data by gating rid of the unsubstantial data.

3. U.S. patent 4,150,400, Kwan Y. Weng, Apr. 17, 1979 used a coarse scaning device for analog signals that were encoded and compressed to distinguish between the white background and black elements of the character. The invention compacts the data by geting rid of some elements of the data.

4. U.S. patent 4,232,375, John F. Paugstat, Nov. 4, 1980 used to generate a compressed data record for the type of merchandise transaction being processed. The invention compacts the data by using specified codes for the limited groups of data.

5. U.S. patemt 4,282,546, Glenn A. Reitmeier, Aug. 4, 1981 used compreasing the television image size using in terpolated pixel values and geting rid of some unsubstantial data.

6. U.S., Reference Manual-IBM 1401 Data Processing System, Apr., 1982. The IBM System used zero-suppression operations in order to compact the data.

In all examples above the previous art was based on previous theory. Thus, the development of coding systems in meaning of compressing the data was given in the mathematical theory of information, particularly by C. Shannem, where it was shown that the speed of transmission cannot be better then a technical speed of a channel of transmission, that means the impossibility to compress the data but only to improve the quality of coding, as was shown in examples above. Particularly for binary systems the best speed is: speed/3.32...(numbers per second), that is very close to speed/4 for the hexadecimal numerical system.

According to the above -mentioned statement of the prior theory and shewn examples, the prior art in significant compression of a data do not exist.

As for embodiments of block diagrams, that are shown on the drawings, the similarity in appearance of previous block diagrams that might occur would not represent the process of the present invention, as far as it was not known the compressing process that is invented by the auther of the present invention.

The author of the present invention challenged the base of the theory of information and made a discovery, that contradiet to an impossibility to compress a data beyend speed of the channel of transmission and trivial at new practice, and he based on it the novelty of the invention.

As far as the now mathematical conception, that was ereated by auther, was never published, it is described and explained below. The terminology and the symbels that are used below and partially in section "description of the profered embodiment" are not eartainly using the computer language, but rather they using language of mathematics. Only the invention itself are using the language to enable any person skilled in the art to make and to use it.

So, the new mathematical conception consist of: If it could be found out once a process that decrease a number of bits, or a number of any other representation of the data, at least (I + a) times, where (a) is any number or fraction, it would be possible to increase the capicity of memory, or the speed of the transmission channel, as many times (n) as ordered, by repeating the process (m) times, so that: n = (I + a)^m.

However the theory of information examine only mathematical structures of a certain type, such as the sum: k_I*p^x+...+k_{x +I}*p^o as well as other methods and combinations, thet prove that the best speed of transmission is log_rn, that makes it impossible to increase the speed of transmission even (I + a) times.

It should be used some strenger structures and processes. The author of the present invention uses a coding process that uses the dimentional co-ordinate of the plase of the data as it value, as described in this section and the section, entitled "description of the preferred embodiment". This process allows to increase the speed of the transmission (I + a) times:

Let's take a group of bytes:

where (b) is a number of bits in each byte, so it can represent (n) numbers from (0) to (2^b), or log_Io (2^b*n) digits.

It can be changed into:

where each dark area is the bit in the state "on" (or "off"), and each blank is the group of one or more bits in the state "off" (or "on"), that determine the distanse _dk. At least for some d_k, the distance is big enough to be coded by the process that representing d_k as new group of bits, and it will occupy only lg₂(d_k) bits.

The probable amount of such giving in bits, for a big enough amount of data, is a 75% of all bits, as will be shown in the section entitled "description of the preferred embodiment".

So, it can be coded (I + a) bits of information for each bit for big enough amount of data, that contradiet to the theory of information and break through it limitation.

The practical algorithm of the coding process is described in the section "description of the preferred embodiment".

DISCLOSURE OF THE INVENTION

The present invention break through oapicity limitation of memory and give also increasing of speed of computers and related devices. It allows to build (or suply with necessary improvement) less expensive computers, computer systems and related devices with practically unlimited memory eapieity for most of existent problems, using with invention existent at now technological level in computers that will give also a possibility to have instead of main storage and auxiliary storage more compact representation of them in computer system that allows also to use less expensive and portable memory in many traditional and new areas. Last object will increase the processing speed, or at least compensate coding and decoding time, because of absence of auxillary devices.

Other object of the invention is the possibility to use new coding and decoding process for transmission and reception in communication and information systems, that will increase in many times the density of transmission existent at now.

BRIEF DESCRIPTION OF THE DRAWING

Thus present invention will become more readily in greater detail hereinafter relative to non-limitative embodiment of the invention and the attached drawings, wherein show first drawing as block diagrams of systems and devices that may be build on principles of tho invention or nay be used with it and useful in explaining these principles and next drawings are block diagrams of one of useful variation of the process, so:

FIG. 1A is a block diagram of coding system that may be used with the invention.

FIG. 1B is a block diagram of decoding system that may be used with the invention.

FIG. 1C is a bloek diagram of control unit for repeated codig systems that may be used with the invention.

FIG. 2A is a block diagram of one of useful variation of the coding process that may be used with the invention.

FIG. 2B is a block diagram of one of useful variation of the decoding process that may be used with the invention. FIG. 2C is a block diagram of one of useful variation of the repeated coding process for coding and decoding that may be ussd with the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENT

Hereinafter, for illustrative purposes, only the one example of each type of systems and processes is described without however limiting the scope of the invention thereto.

FIG. 1A diagramatically illustrates an arrangement of a coding system that codes some data and it is compressing it into less number of bits, using the coding process that uses the co-ordinate of the place of the "on" (or "off") bit as the new value of the data.

FIG. 2A illustrates the compressing coding process that may be used as following:

Step one (25) consist of groupping the original bytes in one set of numbers, big enough to make work the probability of average data and to give may compression. It can group for instance a thousand bits or more for one coding. If the process of the coding will be repeated, the bigger group of bytes has to be taken. If different arrangement of the process was choosen, this step can be emited.

Step two (26) consist of enlarging of the original bytes. This step is used only if different base of the original bytes from the choosen base is represented, in order to have the number of bits in the group as choosen.

Stop three (27) may be used as following:

Let's choose as original byte any number of bits. The group of choosen bytes is changed into a set of bits. The distances between each of two "on" (or "off") states of nearest bits will put together a new set of numbers. For big enough amount of data the probability to find a distance of (0) is 25% of all bits. Accordingly for each next distance: (1, 2, 3, ..., n) the probability will be

12.5%, 6.25%, ..., (1/(2^{n + 2}))*100%. General case can be as any correlation of distances. This step may consist of following substeps:

1. From the last set of distances between nea rest states of "on" (or "off") bits the new set is choosen. It represents only distances (O) and (I), so that the first set of bits is coded into 57.5% of bits plus several bits for identification of the length of the new set, coded separately, that will occupy only log₂L + e, where (L) is the length of the new set and (e) is some number of bits for additional coding. This subset can be defined using any of the known computer processings or by other means,and it can be coded as following:

From the original set of distances it is choosen one subset that includes distances bigger then (O) and (I), and second subset that shows thir place in the original set of distances. The left numbers are coded into memory as subset of distances (O) and (I), choosen from the original set, with identification of the beginning and the end of the subset in the memory, as was described above.

2. the two subsets, that were drawn out, are coded into memory the same way, that was coded the original set of distances in the substep before. The process could be continued at least until it is economical to represemt the last of lefted subsets as subset of distances coded directly. It can be the second subset, that already is economical enough. In that case it will give 5 subsets and 5 specified bytes to identify the length of each subset:

(1) (2) co-ord. of distances of secend subset

(3) from second subset (4)

(5) co-ord. of distances

of subset (4)

3. All subsets that consist of more then two different numbers (subset (4) in the last example) are coded directly, using for economy bytes with different number of bits, for instance:

Subsets (2) and (5) are added to set of original distances of the next process of compressing on the same data.

Step four (23) consist of coding each of the beginnings and ends of coded into memory subsets as distances B_{i +I} B_i, as was already found as the length of each subset above, by giving an address to this specified bytes and subsets and placing them in memory in known order.

Step five (29) consist of rowriting the coded data into memory as group of bytes, original for using computer system, if it is the last process of compressing on the same data. This step might be emited if different arrangement of the coding process was choosen.

FIG. 1B diagrammatically illustrates an arrangement of decoding system that decodes the coded data to original form for using computer system.

FIG. 2B illustrates the decoding process for data that was coded by the coding process above, and may be used as following:

Step one (30) consist of groupping the data in shape of original bytes to one set of bits, as it was coded in the coding process.

Step two (31) consist of decoding the beginnings and ends of each of the coded subsets in precise order.

Step three (32) consist of decoding each of subsets, to create the original set of numbers that was before the coding, using any known computer processing, from subsets of distances and its values. For instance, the 5 subsets from the last example of coding, can be used to reconstruct the original set of bits.

Step four (33) converts the decoded set of bits to original for using computer system form of bytes, if it was required in choosen arrangement of the process. FIG. IC diagrammatically illustrates an arrangement for control unit for repeated process of the codings and decodings.

FIG. 2C illustrates an repeated process of the codings and decodings, and may be used as following:

Steps (34), (35), (36), (37), (38) and steps (40), (41), (42), (43) are the same as the steps (25), (26), (27), (28), (29) and the steps (30), (31), (32), (33) respectively. The ordered level of repeated compressing the data (39) order the number of codings and decodings, using a control unit, depending on the volume of the memory that is occupied by the data after each of the codings. The number of the codings on the same data is written into specified byte of the memory. The process can be repeated as many times as necessary to receive desirable compression of a data. In order to avoid the possible mistake during the coding it can be used the decoding process and it comparison with the original data.

It may be involved equipment FIG. IA, FIG. IB, as sending units (I), (6) and as receiving units (5), (10), that are in infermation, communication, computer and related systems and devices. It also can be created separated data processing and compressing systems and devices, including software and hardware that are based on the coding and decoding precessis above. Such systems and devices can use different level of compression the data for different information, depending on ordered accessibility of information of different priority of the coded data. It woud be a "multilevel depth process", as far as it requires more time for a higher level of compression of the data. So, the more compressed data will be deeper in the memory amd will requires more time for access. The accessebility of the data can be increased by the coding the data tree, involving with each level of the coding new coded blocks, In case of decoding that information, the minimum access time will be needed for each block of information.

Claims

THE CLAIMS

1. A process for coding and compressing a data using the co-ordinate of the "on" (or "off") state of the bit as the new value of the data; any information can be represented as combination of "on" and "off" states of bits ef memory, then it can be determine the distances between "on" (or "off") states of bits, or just "points" (or "blanks"), and such distances can be coded in a new group of information, software and hardware for such process and data storage and processing apparatus.

2. A coding and decoding processes in accordancewith claim I wherein; from original set of distances it is choosen one subset expeling distances (O) and (I), and second subset that shows the place of that distances in the original set, the left numbers are coded into memory as a subset of distances (O) and (I), choosen from original set, with identification of the beginning and end of the last subset in the memory, the two subsets that were drawn out before are coded into memory the same way that was coded the original set for distances (O) and (I) and the process is continuing at least until it is economical to represent the last of the lefted subsets as subset of distances coded directly; from the set of coded data the bytes that identified the beginnings and the ends of each of the coded subsets are decoded in precise order, according to each of decoded subsets, it is created the original set of numbers as before the coding, that process can be al so used to compare the coding data to original.

3. A process of coding and decoding in accordance with claim I wherein the whole process can be repeatedas many times as nesessary according to the desired level of compression of the data, according with accessebility and volume of data, using the coded numbers again as a new data for compression.

4. A data storage and a processing apparatus inaccordance with claim I wherein: the systems and devices for repeated compressing of the data using a control unit that order the number of codings and decodings depending on the volume of the information and priority of accessibility of the information; the control unit that also uses in a control uiit a different level of compression the data for information of different priority and accessebility.