Mobile Agent-based Secure Cloud Data Center Exploration for Load Data Retrieval Using Graph Theory

This paper addresses the load information collection problem for load balancing the cloud data center. This work models Cloud data center as a graph with vertices denoting the servers hosting Virtual Machines and the edges corresponding to communication links among the servers. As Virtual Machines are created and released over time, a load balancer must keep track the load of the servers in cloud data center in order to distribute them uniform among the servers so as to have a load balanced cloud data center. This work harnesses mobile agent concept in cloud data center for load information collection, since both the mobile agent and cloud computing technologies are promising and commercially useful. The idea is to securely explore the cloud data center network quickly with mobile agents to collect load information from the servers and reporting them to load balancer as fast as possible. The goal is to minimize the cover time of the network and minimize the space requirement during load data collection. This paper proposes a secure network exploration algorithm for load data collection that decreases the time taken for exploration and space requirement. The theoretical analysis shows that the proposed approach takes O(logdn) time for network exploration, where as other deterministic approaches used for comparison take more time.