원문정보
초록
영어
With rapidly growing influence and demand for cloud computing services, issues involving control, migration, security, availability of data, trust matters directly related to sensitive information stored can no longer be treated lightly. Through our work as explained in this paper, we discuss these topics, and also review several algorithms which are used in cloud computing. These algorithms are implemented in a cloud computing architecture to balance load to improve efficiency. Weighted active monitoring load balancing algorithm, dynamic load balancing algorithm, static algorithms, and ant colony algorithm are all used for load balancing in distributed computing systems. In this paper, we propose a load balancing technique based on Tree Parity Machine. Our approach is to design a Tree Parity Machine based model that would distribute workload evenly among all nodes. The technique proposed by us is not complicated. It has been conceived such that it can work efficiently with training sets that have been proven to produce intended results. Our Tree Parity Machine based model would predict the demand and, accordingly, allocate necessary resources. Thus, it would be able to maintain active servers to meet existing demands. As a result, consumption of energy can be significantly lowered. The conservative approach of over-provisioning is, thus, avoided.
목차
1. Introduction
2. Load Balancing in Cloud Computing
3. Genetic Algorithm
4. Tree Parity Machine Implementation
4.1. Generation of Queries
4.2. The TPM Model
4.3. Protocol
5. The Proposed Load Balancing Algorithm with TPM
6. Algorithm
7. Result
7.1. Connecting links
7.2. Adder
7.3. Activation Function
8. Conclusion
Acknowledgment
Reference