Proposed Framework for Fog Computing to Improve Quality-of-Service in IoT Applications

In this era of IoT, edge devices generate gigantic data every second. The main aim of these IoT networks is to infer some meaningful information from the collected data. For the same, the huge data is transmitted to cloud which is highly expensive and time consuming. This huge cost is significantly reduced with introduction of Fog Computing (FC) which suggests performing data processing closer to its generation site. FC suggests preprocessing enormous data ahead of forwarding it to cloud by introducing a virtual layer between IoT and cloud, viz., Fog layer and thus accomplishes several benefits like reduced latency, low communication cost, reliability, and scalability. These benefits strongly advocate its employment in real-time application. However, FC also bears some challenges despite several benefits. First and foremost, the processing capability and storage at fog layer is limited in contrast to cloud. Hence, rigorous research is taking place in the direction of devising effective and efficient framework to garner utmost advantage of introducing fog layer. Here, in this chapter, we propose a framework that aims to improve QoS (Quality-of-Service) by providing reduced latency and load balancing at fog layer. This improvement in QoS is achieved with help of data aggregation and load balancing. In the proposed framework, an overburdened fog node requests its neighboring node to share its load. Additionally, it suggests implementing various techniques to aggregate data ahead of transmission. Resultantly, the proposed approach improves QoS by outperforming the existing approaches by preventing bottleneck in the network.