Simulation of Resolution of CS Problem for Multiple Common Variables in Multiprocessor Environment

Abstract

The high speed computing requires multitasking and parallel processing to increase the throughput of the system and optimizes use of system resources. In the multitasking environment, when a great number of processes execute in parallel, there may be some variables, which are common or shared between different numbers of processes. Due to this sharing, different critical sections are produced. The research work presented in this paper resolves the critical section (CS) problem of multiple common variables using multiprocessor environment. The development of a simulator for the resolution of CS problem has been presented. It is a comprehensive tool, which runs a simulation in real time and generates useful data for evaluation. A user-friendly and mouse-driven GUI has also been integrated. The developed system has been put through extensive experimentation. Results are taken using different sets of processes for different number of common variables on a number of processors. The evaluation results are very promising and could be used to further enhance performance of multi-user and multitasking operating systems running under different processors to deal with common variables.