MIST: Fog-based data analytics scheme with cost-efficient resource provisioning for IoT crowdsensing applications

Development of Internet of things (IoT) has revitalized the feature scale of wearables and smart home/city applications. The landscape of these applications encountering big data needs to be replotted on cloud instead of solely relying on limited storage and computational resources of small devices. However, with the rapid increase in the number of Internet-connected devices, the increased demand of real-time, low-latency services is proving to be challenging for the traditional cloud computing framework. Although, fog computing, by providing elastic resources and services to end users at the edge of network, emerges as a promising solution, but the upsurge of novel social applications such as crowd sensing has fostered the need for scalable cost-efficient platforms that can enable distributed data analytics, while optimizing the allocation of resources and minimizing the response time. Following the existing trends, we are motivated to propose a fog computing based scheme, called MIST (i.e. a cloud near the earth's surface with lesser density than fog), to support crowd sensing applications in the context of IoT. For cost-efficient provisioning limited resources, we also jointly investigate data consumer association, task distribution, and virtual machine placement towards MIST. We first formulate the problem into a mixed-integer non-linear program (MINLP) and then linearise it into a mixed integer linear program (MILP). A comprehensive evaluation of MIST is performed by consideration of real world parameters of the Tehran province, the capital of Iran. Results show that as the number of applications demanding real-time service increases, the MIST fog-based scheme outperforms traditional cloud computing.