The document describes a Cluster Scoring Based Task Scheduling (CSBTS) algorithm for scheduling real-time tasks in a virtualized cloud environment. The algorithm aims to decrease task completion times and save energy by dynamically adjusting VM creation, migration and cancellation. It proposes calculating a cluster score based on average transmission power and computing power to assign tasks. The cluster with the highest score is selected. Local and global updates are used to adjust scores based on latest resource status. The approach considers storage capacity and splits jobs into subtasks assigned across multiple clusters to improve performance and energy efficiency.