An improved Harmony Search algorithm with group technology model for scheduling workflows in cloud environment

Cloud computing is a latest innovation in technology in which provide services like infrastructure and applications to end-users as per pay-as-you-go model. It can provide virtualized services according to the requirements of end-users varying with time. The resources of the cloud are available everywhere and user can access these resources anytime. Cloud computing provides virtualization of physical machine into several virtual machines, which increases the resource availability. It provides high performance to the users, and profit to the service providers of cloud. Cloud resources are allocated dynamically on demand of users and shared by multiple numbers of users. The challenge is to optimally map the resources to users in order to satisfy users QoS constraints, known as task scheduling. Due to NP-complete nature of scheduling problem, exact solutions cannot be found in finite time. Therefore, many researchers apply heuristic or random search techniques as Particle swarm optimization (PSO), Ant colony optimization (ACO), Genetic algorithm (GA), Cuckoo Search (CS), Min-Min algorithm etc. to get solutions of such problems which are near optimal. We continuing our research in that direction too and we are proposing a workflow scheduling algorithm for cloud environment based upon Harmony Search algorithm merged with Group technology aiming to reduce makespan. We compared our harmony search based algorithm with Particle swarm optimization (PSO) and Genetic algorithm (GA) and Min-Min scheduling algorithm. The experimental results show that our proposed algorithm provides better results in terms of execution time.