An energy efficient resource allocation for heterogeneous fog network

Abstract

Today, the mushrooming of IoT devices are indispensable. It is newlineapplied in all technologies and huge quantity of data are produced by such newlineInternet of Things (IoT) devices. These data are typically handled by cloud newlinecomputing services. However, it may cause problems associated with newlinemaximum latency and system failure. A technique called latency aware newlinecomputation along with fog virtualization was established to encounter the newlineabove problem. But this solution was hampered by limited resources and newlinedelay in job scheduling and resource management in fog. In virtualization, newlinethere is a provision to create an individual ambience for each device. This will newlineshield them from the impact of resource battle among devices that share the newlinesame fog nodes in the same pool of resources. However, VMs are too newlinecomplicated for time sensitive work. As a result, fog computing is newlinerecommended as an alternative to virtual machines (VMs). Though virtual newlinemachines may take only a few seconds to boot up, even this short amount of newlinetime is too long for delay-sensitive jobs and the budget of VM monitoring is newlineprohibitively expensive. But the performance suffers greatly due to increase newlinein VM. Therefore, container is adapted to address the above problems. It is newlinelightweight when compared to VM. A Scheduling model called CBSTA newline(Container Based Task Scheduling Algorithm) was desired for task newlinescheduling in fog. The task execution procedure happens in two stages: newlinedeciding whether to accept or reject the job, and then allocating the approved newlinetasks for execution. In the proposed algorithm it is considered that the newlineapproved tasks are done within delay limit. In this research work these newlineunsolved issues in the proposed algorithm are studied in detail to identify the newlineproblems not yet solved and provide a novel solution for the identified newlineproblems efficiently. newline newline