مقاله انگلیسی رایگان در مورد فعالسازی هوش توزیع شده با کمک کنترل کننده IoT آینده – الزویر 2018

Future Internet is expected to be driven by the prevalence of Internet of Things (IoT) where it is envisioned that anything can be connected [1]. The hype around IoT is that it is the next technological revolution of the current world [2] where hundreds of billions of things will be interconnected. IoT has started to shape into reality from its hype by and large due to recent advancements in ubiquitous technologies such as Radio Frequency Identification (RFID)/Near Field Communication (NFC), Wireless Personal Area Network (WPAN), high speed communication (4G/5G), Bluetooth Low Energy (BLE), etc. Advanced developments in the sensing and actuating technologies also contribute to the rise of the IoT popularity. This rise in connected things has already taken its number beyond current world’s population and expected to impact every aspect of human life. Currently, there are almost two connected things for every human. The ratio is expected only to accelerate in the coming days. The challenge of collecting and sharing the context information (ConIn) from these connected things has been addressed in earlier research [3–8]. The challenge has been addressed by architecting IoT platforms via mostly middleware solutions. Each middleware solution addresses different IoT challenges; for example, device management, context information collection and sharing, context-awareness, interoperability, etc. [4]. However, there is no single middleware solution or IoT platform that solves all these IoT challenges. An ideal IoT platform capable of providing solutions to all IoT aspects has not yet been designed [4]. Furthermore, most of the IoT platforms solutions are cloud centric [3–5,8]; recently Cisco coined the term fog computing, i.e. edge computing closer to the actual devices [9]. Lately resource constrained devices such as SmartDevices and raspberry pi have enriched in computational capabilities and at the same time price has become more affordable. These devices have the potential to be exploited as IoT gateways and have already been demonstrated in earlier research [10,11]. Emergence of these devices paves the way for computing at the edge of connected things, for example, raspberry pi (Raspberry Pi 3 is a credit card-sized computer with 1.2 GHz quad-core CPU, 1 GB RAM, built-in support for BLE and Wireless LAN, and priced at US $35 [12]) can be employed as an IoT gateway for smart home, mHealth, smart farming, factory automation, shipping, etc. Most of earlier research more or less agree on the typical three-layer IoT architecture as depicted in Fig. 1 [3–5,13]. It shows IoT application where things are connected to a gateway locally and gateway then collects and forwards data to the cloud for further processing which is also shown in the industry for example by AWS IoT [14]. This brings high latency and bandwidth requirements. However, IoT necessitates latency, i.e. response time as low as possible. When IoT was first coined by Kevin Ashton back in 1999, it was proposed in the context of supply chain management [15]. Due to the technological advancements and progressions in the research within IoT have evolved its vision and transformed the way Internetenabled things are being utilized. IoT vision has been expanded to many other application domains. Such expansion of scopes drives IoT on the verge of experiencing a paradigm shift towards enabling Internet of Everything (IoE) [16]. The focus of which is the integration of people, things, services, context information as seen in Fig. 2. To counter this paradigm shift, new approaches are mandated to research into.