Development of Some Load Distribution Algorithm in Distributed Processing System

Abstract

First Chapter of the thesis is devoted to the basic know, how and development stages of DPS includes an introduction part and the techniques available for Load Distribution, with their performances, approached comparatively. A full survey of different work allocation approaches available in the literature is reported in Chapter 2. Chapter 3 aims at developing a task allocation policy for balance utilization of available resources in distributed computing environment. The purpose of the problem is to set a new processor with a set of m module to obtain an optimum consumption of available processors. We would be assigned to reduce execution costs and the odds of reducing communication and make the best use of the total throughput and the system has been adjusted well to load all the processors that this way in control. In Chapter 4, we presented a fuzzy based clustering method for performance emulation distributed processing system. The membership function is formulated which lies between 0 1, then we categorized the processor and task in four categories, based on their membership values by using the MATLAB, MAMDANI inference engine. In Chapter 5, Provide a system consisting of a set of processors connected with the router and switch. Thus, whatever the complexity of a distributed system must be a reliable transmission. The genuineness of the system is established through its simulation on Boson Netsim followed by FTA for reliability evaluation. In Chapter 6, using locally optimal choice method for the task allocation model in distributed computing. The current problem reduces the total process time of a distributed system in excess of allocation of the jobs of many processors of the system. The algorithm based on a start time and end of task on each processor, which represents the ideal computation time of execution of tasks on processors and that affected by the ITCC. Chapter 7 concludes the thesis and propose other issues which require further investigation and potential research expansion. newline newline newline