An Optimized Task Scheduling Algorithm In Multi Processor By Using Cuckoo Search
Cloud computing is an evolutionary field and has a lot of opportunities in the field of scheduling and localizations of the nodes. Parallel computing is an essential part of cloud computing paradigm as without multiprocessing, a computing environment like cloud can't deliver efficient results. Parallel computing has a wide range of possibilities in terms of scheduling the nodes. Some of these possibilities are named as Directed Acyclic Graph(DAG) and Heterogeneous Earliest Finish Time (HEFT). The algorithms utilize the concepts of Earliest Start Time(EST) which refers to the time when a job starts on a processor and Earliest Finish Time(EFT) which refers to the time when a job completes on a processor. This architecture faces a serious issue termed as time slack. The proposed work aims to optimize the time slack issue by using Cuckoo Search. The evaluation has been done on the basis of Computation Cost Ratio (CCR) and Schedule Length Runtime (SLR).
Keywords - Parallel computing, DAG scheduling, List scheduling, Optimization.