دانلود رایگان مقاله انگلیسی تامین منابع برای خدمات اینترنت اشیا در مه - نشریه IEEE

Abstract

The advent of the Internet of Things (IoT) leads to the pervasion of business and private spaces with ubiquitous, networked computing devices. These devices do not simply act as sensors, but feature computational, storage, and networking resources. These resources are close to the edge of the network, and it is a promising approach to exploit them in order to execute IoT services. This concept is known as fog computing. Despite existing theoretical foundations, the adoption of fog computing is still at its very beginning. Especially, there is a lack of approaches for the leasing and releasing of resources. To resolve this shortcoming, we present a conceptual framework for fog resource provisioning. We formalize an optimization problem which is able to take into account existing resources in fog/IoT landscapes. The goal of this optimization problem is to provide delay-sensitive utilization of available fog-based computational resources. We evaluate the resource provisioning model to show the benefits of our contributions. Our results show a decrease in delays of up to 39% compared to a baseline approach, yielding shorter round-trip times and makespans.

VI. CONCLUSIONS

The centralized processing of multiplicative IoT data in the cloud incurs high delays and accordingly low speed of data processing which are unfavorable for IoT applications and services. Fog computing promises to solve this problem by utilizing available computational, storage, and networking resources for the enactment of IoT services close to the edge of the network. Currently, the uptake of fog computing is still at its very beginning, thus there is a lack of theoretical and practical foundations for fog resource provisioning.

After having motivated our work, we discussed an architecture for fog computing framework, and derived a system model for fog resource provisioning. To evaluate the efficiency of the proposed approach, we simulated the envisioned architecture. We showed that the system model combined with the timeshared provisioning of fog cells for services along with spaceshared provisioning inside services for task requests decreased delays by 39% and yielded shorter round-trip times and makespans.

In our future work, we aim to implement a real-world testbed based on the proposed architecture and to improve the system model for resource provisioning. The architecture can be enhanced by fault tolerance mechanisms, and by adding QoS constraints to the task requests, e.g., deadlines. Another aspect of our future work is the systematic evaluation of a fog landscape to obtain real-world network data for evaluations, e.g., delays and bandwidth.