DEVELOPMENT OF METHODS AND ALGORITHMS FOR COMPRESSING INFORMATION DATA OF TABULAR STRUCTURES

Abstract

The properties of the non-positional number system in the system of residual classes (SRC) are considered: independence of residues, equality of residues and low-bit residuals. Use of the main properties of SRC gives the chance to create effective methods of control, diagnostics and correction of errors of data at introduction of the minimum information and time redundancy, and construction of fault-tolerant and high-speed computer systems of data processing (CSDP) of real time. The research of influence of the basic properties of SRC on architecture and principles of functioning of modern CSDP is considered. The main property of SRC, namely low-bit residuals allows you to use tabular methods of arithmetic operations. In this case, most arithmetic operations are performed in one cycle, which dramatically increases the speed of rational operations. At the same time, tabular methods of using arithmetic operations allow to create on the basis of matrix schemes highly reliable computing devices of computer data processing systems. When applying the methods of special coding of information in SRC, the purpose of which is to reduce the tables of RAM that implement tabular operations. The amount of equipment in the tabular construction of the operating device in the SRC is reduced by at least twice in comparison with analogues implemented on an element basis using a binary positional number system. Also, when performing a modular operation, it is possible to further reduce the equipment by organizing not a single table (RAM), which implements the result of the operation in machine code, but much smaller tables that give answers for each of the k bits, where k – is the bit size required for storing the figure on the basis under consideration. Based on the research data, methods and algorithms for compressing information data of tabular structures in SRC were developed