Improved load balancing technique For real time video processing in Cloud environment

Abstract

In cloud computing, the two primary mechanisms are loadbalancing newlineand scheduling. The uniform workload balance with the available newlinecomputational resources is vital for achieving load imbalance, makes span, newlineand optimum resource utilization. The different types of workloads such as newlineVirtual machines, containers, databases, Hadoop nodes, and applications are newlinedefined as workload heterogeneity. The computing resources are newlineheterogeneous in memory/cache and clock speed hierarchy. The computing newlineresources and workload heterogeneity that leads to optimization problems are newlineNP-hard. Generally, the standalone surveillance system failed to balance the newlinehuge data volume. Handling huge surveillance video and traffic management newlineis more complex. To overcome this, we enhanced the Apache Hadoop clusters newlinesuch as Map-reduce based system for quickening the complicated video newlineprocessing. The input video data is structured and reduced according to the newlinevideo content. A Novel Awaiting Computation Time (NACT) mechanism is newlineproposed for balancing the load in heterogeneous multicore systems. This newlineproposed approach integrates heuristic models by creating expected newlinecomputing time-based workload priorities newline