Load Prediction Algorithm for Dynamic Resource Allocation

This paper presents dynamic resource allocation in cloud computing using load prediction algorithm. Cloud computing is a technology that is of increasing demand these days. Cloud computing provides various kinds of services to the users. In cloud computing the provider dynamically allocate the resources. Doing so, the service provider should have some knowledge about the future resource needs. They can be calculated using the load prediction algorithms. In this paper an algorithm named Enhanced Exponentially Weighted Moving Average (EEWMA) is used to predict the load, which is efficient in terms of both overload avoidance and green computing. EEWMA algorithm predicts the future need of resources effectively and it is suitable for both increasing and decreasing need of resources. This proposed load prediction will be well suitable for overload avoidance applications.