Fog Computing Architecture, Evaluation, and Future Research Directions

IoT has been gaining a lot of attention in recent years. In fact, the number of connected devices has already surpassed the total population on Earth. Current developments in various technologies have paved the way for many IoT services that are now being deployed in different sectors. IoT can extend its scope and service provisioning capabilities with the integration of the cloud computing paradigm. Similarly, cloud cannot only access underlying IoT nodes and provide them with cloud services, but can also create further enhanced services based on the data collected from IoT nodes. I lowever, loT and sensor networks on the ground are often far from the cloud, which is normally accessible via the Internet. Some services require fast response and a great deal of pre-processing and filtering, and may also have security and privacy requirements. For all such cases, middleware, also known as fog, is required between the cloud and IoT devices. We present cloud-IoT integration issues followed by a comparison between fog and cloud computing. We evaluate the performance of fog computing using performance metrics such as processing delay, processing costs, and processing power, and derive the performance gains obtained in comparison to a cloud computing only approach. Finally, we identify sonic future research directions for fog computing.