More Videos...
 

Non-recursive inorder traversal on constructed threaded K-D tree for efficient cloud based space partitioning

Non-recursive inorder traversal on constructed threaded K-D tree for efficient cloud based space partitioning Cloud computing system comprises of a set of cloud nodes scattered over a large geographical surface. These cloud nodes usually helps to handle efficiently a new arrived job in cloud system. Now a day’s researchers used to work for logically partitioning the earth surface such that every partition contains at least one cloud node which will be prime dedicated for handling jobs arrived in that partition. This phenomenon of partitioning the cloud system is called as ‘Cloud Division Rule’. Using spatial data structure we have recursively broken down the earth surface in small zone, namely cloud cluster, containing single cloud node. For getting nodes’ information (like load status, network bandwidth etc.) from the master cloud node traversing all the child cloud node is one of the way. Traversal of eachcloud node is recursively performed in order to provide some services to users’ request in consideredcloud system. As millions of cloud nodes over the earth surface are to be clustered and recursively traversed, so a huge number of stack space and stack operations are internally initiated. To optimize these operations we have constructed a threaded k-d tree and do non-recursive inorder traversal over it. This novel approach makes the searching of ancestor and descendant cloud nodes from the currentcloud node very easy and also optimizes the computation time and internal memory space of the computation unit which is one of the prime aspect for today’s cloud computing Infrastructure as a Service (IaaS).

Recent Projects

More +