SECURE AND EFFICIENT FAULT NODE DETECTION IN WIRELESS SENSOR NETWORKS
Abstract
Propose an included, energy efficient, resource allocation framework for overcommitted clouds. The concord makes massive energy investments by 1) minimizing Physical Machine overload occurrences via virtual machine resource usage monitoring and prophecy, and 2) reducing the number of active PMs via efficient VM relocation and residency. Using real Google data consisting of a 29 day traces collected from a crowd together contain more than 12K PMs, we show that our proposed framework outperforms existing overload avoidance techniques and prior VM migration strategies by plummeting the number of unexpected overloads, minimizing migration overhead, increasing resource utilization, and reducing cloud energy consumption.