Cloud Computing to Fog Computing: A Paradigm Shift

*Syed Rizwan Hassan and Muhammad Rashad*

### **Abstract**

Fog computing scatters the resources throughout the system to provide services close to the edge of the network. This chapter provides an overview of different segments associated with the fog computing paradigm for implementing efficient Internet of Things (IoT) applications. Section 1 provides an overview and motivation behind the provision of healthcare services using cloud and fog computing paradigms. Section 2 provides the literature and research work related to the deployment of healthcare applications using cloud and fog computing architectures. Section 3 provides the architectural design of a fog computing-based remote pain monitoring application. Section 4 provides the simulation parameters and architecture that are arranged for the evaluation of the proposed policy. Finally, Section 5 concludes and discusses the results of simulations obtained on different scales.

**Keywords:** e-healthcare, fog computing, cloud computing, remote pain monitoring, latency, network consumption

### **1. Introduction**

Due to recent technical advances in the field of information and technology, the provision of online services in every area of life has become possible. The IoT technology links the devices in regular use by humans with the Internet and provides an interconnection between billions of devices throughout the globe. According to Ref. [1], the number of such devices is estimated to reach 12 billion by the year 2021, which will bring an enormous revolution in living aspects and social routines of human lives. The limited processing and storage capacity of IoT devices restricts them from implementing applications involving data of heterogeneous nature or performing tasks involving big data. So, most of the IoT applications are designed on cloud computing architecture that allows resourceful cloud servers to access the data sensed by the IoT devices for processing and storage [2]. This integration of the cloud computing paradigm and IoT technology has numerous benefits allowing the deployment of diverse types of applications with different requirements.

Integration of IoT technology in the healthcare industry is shifting toward the remote provision of health services to patients without the physical intervention of doctors. This will not only accelerate the healthcare process but also provide ease to the patients to get doctor prescriptions and diagnostics required by uploading their

health-related data. Smartphones are a perfect example of IoT devices. Cloud computing offers massive computational resources for the fast processing of diverse and huge amounts of data coming from a large number of patients in healthcare applications [3]. However, latency and network load problems arise in cloud-based IoT healthcare applications when implemented on large scales due to a huge rise in data to be processed [4]. Therefore, it is not viable to implement latency-sensitive healthcare systems in the cloud computing paradigm.

Fog computing term was introduced by CISCO in 2012 [5]. The fog computing paradigm provides the solution to the challenges that evolved in the cloud computing paradigm by the provision of computational and storage resources in a distributed manner near the edge of the network to implement IoT applications. The fog paradigm introduces fog devices having limited version of cloud resources between the IoT devices and the cloud servers, which offers computational supports to edge devices to process heterogeneous types of data generated by IoT devices, which results in less delay and reduced network load. Hence, these formerly mentioned prominent features of the fog computing paradigm make it a better choice for the implementation of IoT healthcare applications.

This chapter briefly describes the beneficial aspects of adopting the fog computing paradigm for the design and implementation of healthcare applications. Initially, a literature review covering different healthcare applications designed on the cloud computing paradigm, their architecture, benefits, and challenges is presented. Later, the architecture of the fog computing paradigm to implement healthcare applications in a distributed manner is presented. Correspondingly, a detailed literature review of different applications designed using the fog computing paradigm is presented. Finally, the design of a remote pain monitoring application using the fog computing paradigm is presented with a comparative analysis between fog and cloud-based implementations to show the effectiveness of the fog computing paradigm.

Internet of Things (IoT) applications are drastically growing to facilitate mankind. Before the adaptation of the fog paradigm, mostly these applications are deployed on cloud-centric architectures. Due to the high demand for IoT applications, cloud computing faces numerous challenges, such as high delay, burdened network bandwidth, poor Internet connectivity, scalability, and high execution cost. To address these challenges, fog paradigm comes into play that extends the cloud resources close to the edge of the network by employing fog devices throughout the network. Fog devices have limited resources for the storage and processing of detected information. To provide real-time response to the end users, some applications are interested in providing computational services near the edge of the network but still, some of the applications are there that require big data analysis that needs to be processed at the cloud server. To implement efficient IoT infrastructure, there is a need of seamless and effective orchestration of resources. Fog devices work as smart gateways between cloud and end devices, providing fog and cloud connectivity. The main objective of this research is to address the critical issues involved in the deployment of IoT applications on cloud and fog computing paradigms. A solution is implemented in this research to improve the efficiency of the applications. The second aim of the research is to investigate the architectural performance of the fog computing paradigm from an application perspective and the design of policies to implement applications in a way to achieve optimal performance. Simulations are executed on multiple scales to evaluate the proposed design. The simulations results confirm the effectiveness of the proposed paradigm in achieving a reduction in delay, network utilization and processing cost at the cloud.
