An efficient framework for data offloading and task allocation in edge based iot networks

Abstract

IoT is one of the technologies that has gained exponential rise by providing newlinemachine to machine (M2M) communication and bringing smart devices closer to newlinethe proximity of end users. Due to exponential rise of smart devices, there is huge newlinedata traffic in cellular networks causing more bandwidth and energy consumption newlineof devices, delay in response to the critical IoT based applications. An alternative newlineto this problem is to perform offloading at newly introduced edge and fog nodes newlineof IoT and further allocate the computing resources efficiently to the offloaded newlinedata for processing using task allocation algorithms. newlineIn the thesis, a Priority aware task Scheduling (PaTS) algorithm has been designed newlinefor sensor networks to schedule priority aware tasks to offload data at edge and newlinecloud servers. The problem is formulated as a multi-objective function and the newlineefficiency of the proposed algorithm is evaluated using the bio-inspired NSGA-2 newlinetechnique. The results obtained are compared with the existing energy aware newlinescheduling (Eas)approach. The overall percentage improvement for average queue newlinedelay, computation time and energy obtained for 200 tasks is 17.2 %, 7.08% and newline11.4% respectively. The results obtained show significant improvement when newlinecompared with the benchmark algorithm to show the effectiveness of the proposed newlinesolution. Similarly comparative results for tasks when increased from 200 to 1000 newlinetasks also show subsequent improvements. newline