Tasks management algorithm for distributed system

Abstract

The term "Distributed Real-Time Computing System” is described whenever number of computers interconnected in some fashion such that a program or procedure utilizes this distributed but combined power and gets executed in Real Time. The term has different meanings with regard to different systems, because processors can be interconnected in many ways for various reasons. In its most general form, the word distribution implies that the processors are fixed in geographically separated locations. Occasionally, the term is also applied to an operating environment using multiple mini-computers not connected with each other with the help of physical communication lines but are connected through satellite. In a Distributed Real Time System (DRTS), the single communication channel share by all the processors for Inter-Processor Communication (IPC). A program whose execution is distributed among several processors in a distributed system has the total processing cost equal to the sum of processors costs and Inter Tasks Communication Cost (ITCC), which are function of the amount of data transmitted. An optimal assignment is a distribution of modules that has lowest total Execution Cost (EC). The model discussed in this paper provides an optimal solution for assigning a set of “m” tasks to a set of “n” processors where m >> n, in such a way that allocated load on all processors is balanced according to the relative speed. Distributed Real-Time System finds extensive applications in the faculties, where large amount of data is to be processed in relatively short period of time, or where Real-Time computations are required. Meteorology, Cryptography, Image Analysis, Signal Processing, Solar and Radar Surveillance, Simulation of VLSI circuits and Industrial process monitoring are areas of such applications. These applications require not only very fast computation speeds but also different strategies involving distributed task allocation systems. In such applications the quality of the output is proportional to the amount of Real-Time computations. In a Distributed Real-Time System the execution of a program may be distributed among several processing elements to reduce the overall system cost by taking the advantage of heterogeneous computational capabilities and other resources within the system. For the optimal utilization of processors systematic allocation of task is necessary.