Mohammad Nikravan - Department of Computer Engineering, Islamic Azad University, Shahr Qods Branch, Tehran, Iran

A Distributed Computing System comprising networked heterogeneous processors requires efficient process allocation algorithms to achieve minimum turnaround time and highest possible throughput. To efficiently execute processes on a distributed system, processes must be correctly assigned to processors and determine the execution order of processes so that the overall execution time is minimized. Even when target processors are fully connected and the communication among processors is fast and no dependencies exist among processes the scheduling problem is NP-complete. Complexity of scheduling problem dependent of number of processors, process execution time and the processor network topology. As distributed systems exist in kinds of homogeneous and heterogeneous, in heterogeneous systems the difference between processors leads to different execution time for an individual process on different processors and makes scheduling problem more complex. Our proposed genetic algorithm is applicable for both homogeneous and heterogeneous kinds.

Distributed systems, homogeneous, heterogeneous, scheduling and genetic algorithm