International Journal of Computer Networks and Applications (IJCNA) DOI: 10.22247/ijcna/2019/190369 Volume 6, Issue 6, November – December (2019) ISSN: 2395-0455 ©EverScience Publications 100 SURVEY ARTICLE A Survey on Edge-Based Internet-of-Things Hushmat Amin Kar Department of Electronics and Communication Engineering, National Institute of Technology, Srinagar, India hushmat.kar@gmail.com G.M. Rather Department of Electronics and Communication Engineering, National Institute of Technology, Srinagar, India. ghulammohdrather@yahoo.com Published online: 30 December 2019 Abstract – Internet of Things (IoT) is improving the overall quality of our lives by helping us to connect, measure and control the different parameters of the system in an automated manner. The IoT devices are generating massive volumes of data that needs to be processed and on the basis of the results, decisions are made. The IoT devices have limited resource capabilities, so these devices utilize the services of the cloud servers. The issue in utilizing the services of the cloud is that it fails to provide support for real-time and time-critical applications. In order to reduce the response time of the system, another service layer is added to the architecture i.e. Edge computing. The IoT devices will now send their requests to the edge servers. Utilizing the services of the edge servers will reduce both the network traffic to the cloud and response time of the system. This paper presents a detailed survey of Edge-based IoT taking various parameters like architecture, bandwidth, security, energy, payload, etc. into consideration. Index Terms – IoT, Edge, Cloud, 5G. 1. INTRODUCTION IoT is a revolution in the existing internet where a large number of intelligent devices are connected through the internet. These devices (nodes) sense, gather and communicate the data with each other through improvised communication protocols [1-4]. A large amount of data collected by these devices need processing for extracting intelligence to provide services to end-user. In traditional computing, the data collected by the nodes is uploaded to the cloud server for further processing and results are transferred back to nodes for a needful response. This approach has a drawback of using costly bandwidth and other resources. Also with the increase in data size the transmission time increases which is unacceptable for time-sensitive applications like Smart transportation [5] smart city [6-8], Smart Grid [9,10], Smart healthcare [11]. Battery life is an important concern in an IoT device, so it is better to send the data to a nearby edge device with higher power backup and computational capabilities. The processing of data nearby the source will reduce the transmission time, power cost, etc. The edge device gives the nodes services like processing and storage close to the device at the edge of the network instead of than sending it to the cloud server. Thus, the amount of the data flow is reduced which in turn utilizes lesser bandwidth of the network. It minimizes the response time of computational nodes and also reduces the pressure of traffic and computation from the centralized cloud servers. By utilizing the services of edge nodes the IoT devices having limited battery can shift the processing and communicational overheads to the edge node having more resources as compared to the IoT nodes. Thereby, increasing the overall IoT node life. 2. SURVEY OF IOT AND EDGE This section discusses the fundamental concept of the IoT, EDGE computing and the benefits of combining these two concepts. 2.1 IoT IoT is an interconnected ecosystem of uniquely addressable devices having the capabilities of sensing, computation, and actuation and the ability to communicate and interoperate through the internet. The IoT can be defined as a dynamic infrastructure providing self-identifiable adaptive capabilities in nodes, in order to make them intelligent. These nodes recognize the triggers in the surrounding environment and accordingly react in an appropriate manner. This new environment will create new application services and each application service will achieve a common goal. IoT is an evolving technology that is expanding its horizons in different areas at a very fast rate. The number of IoT devices is expected to increase to 50 billion connected devices, with over 200 billion intermediate connections by 2020 [12]. The main aim of IoT technology as described by Gartner [12] are: • Integration of the physical and virtual worlds. • Embedding the intelligence everywhere. • Effect of the digital transition on technology The improvement in the areas of sensors, Big Data, embedded systems, ubiquitous computing, cloud computing,