CPU scheduling should preserve fairness and avoid processes from do not ever obtain CPU. Modern operating system era faces multitasking on computer operational environment. If CPU scheduling is efficient, high computation could be done correctly and system could be retained stable. One criterion that must be achieved by scheduling algorithm is minimization average waiting time for a set of processes in gaining CPU allocation. There are several algorithms for CPU scheduling; one of them is Round Robin. Round Robin supplies quantum that is same for each of processes. However, there is no standard for quantum. Inevitably, if quantum is very high, response/waiting time for each process could be high, and otherwise, there is an increasing CPU overhead for context switching. This research concerns with improving Round Robin performance. Our approach is to combine Round Robin with Genetic algorithm. In this approach, an individual is a quantum that will be iterated for achieving best quantum that will produce minimal average waiting time. We use integer number for representing a chromosome with length three. Furthermore, we use roulette wheel method for parent selection and steady state replacement technique for survival selection. By using one point crossover and flip mutation, this approach can result better average waiting time than that were found in references used.

Cpu Scheduling Round Robin Algorithm Genetic Algorithm