Abstract-The tremendous growth of Internet of Things (IoT) and due to limitless services provided
by the cloud, the devices connecting to the cloud are increasing exponentially. Due to this, cloud has to process large amount of data resulting in delay due to which quality of service provided decreases. The solution to this problem is Fog/Edge computing. Because of the benefits provided by the Fog/Edge computing the computing pattern has shifted gears
to Fog/Edge computing. Fog computing bolster cloud computing by increasing computational speed thus decreasing latency, immense storage, provides security and privacy and efficient bandwidth usage.
This paper presents the real-world applications of Fog computing/Edge computing and discuss the challenges faced and solutions to the addressed challenges.
Keywords- Cloud computing, Internet of Things (IoT), Fog computing, Edge computing, Latency.
Cloud computing is a service based computing paradigm where customers pays for how much they use, but latency and the bandwidth were not given much importance. Due to exponential growth in the Internet of Things (IoT) devices connecting to the cloud, the cloud computing is inadequate to handle enormous data generated by these devices and this is where Fog computing comes in to play. According to the analyst firm Gartner, 8.4 billion devices were connected in 2017 and 20.4 billion IoT devices to be deployed by 2020 1.As the number of devices increase it becomes more challenging to handle these devices. This increase
causes many problems, delay in processing the data increases, congestion, security related issues. To address these issues we make use of Fog/Edge computing.
In 2010, CISCO introduced the concept of fog computing and revamped the computational and storage methods between cloud and the end users 2. Fog computing is an extension of cloud computing and thus providing enhanced services. The notable features of Fog/Edge computing are :
ii) Very low latency
iii) Services large number of nodes
iv) Covers large distance.
It is called fog computing because fog is closer to ground and the computation between the cloud and end devices happen away from the cloud. Fog/edge computing is used in Smart Transportation, Healthcare etc.
Fog computing provides networking services between cloud computing data centres and end devices, virtually located at the edge of the network.
When compared to loud Fog computing provides better security because data processing takes place at many stages.
Figure 1 3 shows the difference between fog computing and cloud computing architectures.
The main difference between Cloud computing and Fog computing is in cloud computing the data from end devices is directly uploaded to the cloud. Whereas, in fog computing the data is first uploaded to Fog nodes and from fog nodes to the cloud. Fog node is a cloud server with reduced capability from that of actual cloud.
Figure 2. Three Layer Architecture of Fog
The three-layer architecture of the Fog computing is shown in figure 2 4.It consists of three layers.
Device layer: It is also known as End device Layer.
It consists of devices that are connected to the network. The devices can be mobiles, smart vehicles, smart cards, sensors etc. The large amount of data which is generated by these devices are sent to the above layer using WiFi, LTE, Bluetooth etc.
Fog Layer: This layer is near to the end-users and it consists of fog nodes. Fog node is capable of storing and computing the data received from the end devices. These fog nodes are mobile due to this feature the services provided by fog nodes are ubiquitous.
Cloud Layer: This layer consists of servers and
storage devices and provides various services such as VANETs, smart cities, Industry 4.0 etc. Platform-as-a-service (PaaS), Software-as-a-Service (SaaS), and Infrastructure-as-a-Service (IaaS) are the services provided by cloud.
The smart device will connect to the one of the fog using wireless technology. The fog nodes can communicate with other nodes using wireless or wired technology. The fog nodes and cloud are connected using IP network. Fog nodes provides security to the data because the data is not directly processed in the cloud, instead it is processed in the fog node and uploaded to the cloud. But, topology design, distribution of incoming data from the end devices to the fog nodes is complex.
FOG COMPUTING CHARACTERISTICS
Fog computing provides many benefits compared to cloud computing. Some of the essential characteristics of Fog computing are:
• Mobility: It supports communication between mobile end devices.
• Large number of nodes: Fog services many number of devices with low latency.
• Large coverage: The services provided by the fog covers large area.
• Diverseness: Fog can operate any environment and can collect data from various devices.
• Real-time response: Fog provides support for real time applications.
• Low latency: Since fog nodes are near to end devices latency is reduced and processing is very fast.
APPLICATIONS AND FOG COMPUTING EXAMPLES
The fog computing can simultaneously execute many applications by analysing the data received by the end devices and provide real-time response because real-time applications used in daily life needs response very quickly. Some of the real-world applications are described here.
The principles of Mobile ad hoc networks (MANETs) are used in Vehicular ad hoc networks (VANETs) 5. VANET is a wireless network used to exchange data between vehicles. VANET is used in smart transportation, smart parking and traffic management. In VANET vehicles are supplied with On-Board Units (OBU) and Road Side Units (RSU) which provide better traffic management, road safety and comfort to drivers. In VANET, vehicles can communicate with each other so better services are achieved.
OBUs and RSUs sense data from the vehicles and its surrounding and this data is processed at the cloud. But Quality of Service is not assured in VANETs.