Article Preview
Top1. Introduction
IoT was initially proposed in 1999 by Kevin Ashton (Ashton, 2009), since then it has been gaining popularity and benefiting every aspect of human life. The basic purpose of the IoT is to connect everything online, reducing human participation and automating business & private spaces (Botta et al., 2016). In recent years, the IoT has benefited smart cities (smart houses, smart buildings, smart surveillance) (Gharaibeh et al., 2017), smart healthcare (smart wearables, personal monitoring) (Catarinucci et al., 2015), commercial (shopping, retail), industrial automation (particularly production) (Atzori et al., 2010), and even agriculture (Elijah et al., 2018)(Brewster et al., 2017). IoT is becoming increasingly important in today's digital age, resulting in a rapid increase in the number of connected devices (Middleton et al., 2015). Massive amounts of data will be generated by such widely distributed IoT devices at the network's edge. Processing, analysing, storing, and sending this massive amount of data in the centralized cloud is expected to result in increased bandwidth utilization, latency, and network congestion (Shi & Dustdar, 2016)(Shi et al., 2016)(Mahmud et al., 2019).
Furthermore, cloud computing cannot meet the needs of IoT applications that require real-time response, such as fire alarm alerts, healthcare, and self-driving cars, because cloud servers are located far away from IoT end-user devices (Dastjerdi & Buyya, 2016)(Bangui et al., 2018). Edge computing, which delivers services close to IoT end-user’s devices without requiring high-speed Internet connectivity, has become a new paradigm in recent years to decrease network congestion and provide real-time services closer to IoT end-user’s location (Satyanarayanan, 2017)(Buyya & Srirama, 2019)(Ai et al., 2018). Edge computing architecture is displayed in Figure 1, which consists of three layers, namely the device layer, the edge computing layer, and the cloud computing layer.
Figure 1.
Edge computing architecture
Device layer- The device layer includes all IoT devices such as sensors and actuators, security cameras, tablets, and mobile phones etc. All connected devices at this layer have a ubiquitous presence on the Internet.
Edge computing layer- Edge nodes located at the edge computing layer are devices that can direct network traffic and have limited computing power. Edge nodes can range from dedicated hardware, switches, routers, to small data centers that act on raw data. The data produced by the device layer is no longer transferred directly to the cloud computing layer but is instead processed and analyzed locally at the edge computing layer, providing real-time feedback to the end users.
Cloud computing layer- Cloud data centers with nearly infinite resources can belong to different providers, are included in the cloud computing layer. This layer is used for long-term data storage and resource-intensive computation tasks such as training deep learning models and performing big data analysis.