Task scheduling in cloud computing environment using hybrid genetic algorithm and bat algorithm

Cloud computing refers to the delivery of computing services over the internet. It helps users to only pay for their usage. Cloud computing becomes one of the fastest-growing technologies since it can lower costs, higher efficiency, and scalability. Managing files and services on the local storage devices is no longer used with cloud computing that doing the same over the internet. One of the important and challenging components in cloud computing is task scheduling. Task scheduling is an organized user’s task then executed with the suitable resource to perform that service efficiently. Task scheduling problem also categorized as NP-hard problem where optimization technique can be used to solve it. Nowadays, many task allocation techniques are used but the most efficient technique needs to be figured out. Meta-heuristic algorithms are mostly used to solve this problem. For example, Genetic Algorithm with Particle Swarm Optimization, Genetic Algorithm with Artificial Bee Colony Algorithms (ABC) and Genetic Algorithm with Ant Colony Optimization Algorithms. In this thesis study, hybrid Genetic Algorithm and Bat Algorithm proposed to solve the task scheduling problem. Genetic algorithm was widely used because of its accuracy and simplicity. But it will become slower in certain cases that include larger problem size. Hence, Bat Algorithm (BA) can increase efficiency and performance because it provides an efficient scheduling mechanism. BA also will minimize the execution time and deadline. This research will conduct comparison of hybrid Genetic Algorithm and Bat Algorithm (GA-BA) with Genetic Algorithm (GA) and Bat Algorithm (BA). Furthermore, CloudSim simulator will be used to evaluate the performance of this algorithm. The result of the algorithm performance will be appeared in web application system. The features of this system are displaying makespan time for each run of simulation and calculating the average of makespan for all simulations run