Data Placement in the Fog Computing Environment Using Improved Non-Dominated Sorting Genetic Algorithm

With the rapid growth of IoT devices, a huge amount of data is generated, and one of the most important issues in dealing with this data is its proper management. Storing this massive data can lead to a reduction in service quality requirements. Cloud computing also fails in this scalability. Therefore, fog computing can solve the cloud's shortcomings and manage the data produced by IoT devices optimally. One of the important topics in the field of data storage is data replica placement. This issue's significance becomes evident in scenarios like smart hospitals, where data loss equates to jeopardizing patient lives. Consequently, data replicas are essential, and access to these replicas should also consider the QoS requirements. In this paper, a meta-heuristic algorithm based on genetics is presented for placing data replicas on the nearest node to the IoT devices, which aims to simultaneously reduce energy consumption, cost, and access latency. According to the evaluations made by varying requests of the devices, the latency and cost by the proposed method are ۱۷% and ۲۱%, and with the varying number of fog nodes, the latency and cost of ۶% and ۹% improved, and the overall energy consumption has improved ۱۶%.

Keywords:

Fog computing , Data placement , Data management , Data replica , Meta Heuristic Algorithms