Edge computing for Internet of Things: A survey, e-healthcare case study and future direction
Introduction
Last decade has seen the overwhelming growth of the smart sensors, smart actuators, networking technology, and low-power consuming chips (Ray, 2016a, 2016b). Thus, leading to immense increase of data transmission over the internetwork backhaul. This has certainly put an over burden on the existing cloud infrastructure to efficiently handle all such requests in timely fashion (Dinh et al., 2013; Satyanarayanan et al., 2009; Bonomi et al., 2014). Edge along with Internet of Things (IoT) have emerged as a novel ecosystem i.e. edge-IoT ecosystem that has inherently held the demising power and capabilities to other computing paradigms such as, grid computing, cloud computing and fog computing (Atat et al., 2017; Wang et al., 2016; Ray, 2016c). Obviously, current socio-economic scenario has led to foresee such a novel venture that could be able to exaggerate the speed of social-development upon a surprising ceiling. Indeed, other existing computing solutions are good enough to cater all these social requests, they lack mainly in two areas e.g. (i) latency to serve users' request and (ii) heavy load on internetwork backbone. Resulting an ineffective closure-start for the futuristic smart socialization (Ray, 2016d). This problem becomes more when the industrial involvement come into the scene (Ray, 2015a, 2015b).
To solve this problem, edge computing has been proposed in recent past. It observed that the capability of the edge computing and IoT could be merged to create a new genre of ecosystem which would benefit the overall growth of the information and communication technology-based development (Ray, 2014a, 2015c). It is already validated that positioning of network end-devices to near proximity of user/application enhances the speed of response i.e. the network latency is minimized (Ray, 2014b; Wu et al., 2016). Further, the data-intensive applications are readily become efficient when such interventions are deployed in practice. Such capabilities could lead the edge-IoT ecosystem to outperform other existing computing paradigms. Moreover, placing the IoT solutions along with the edge computing provides an excellent opportunity to serve localized smart applications. Doing so in turn reduces the burden of data propagation through the network backhaul (Satyanarayanan et al., 2015; Premsankar et al., 2018; Mao et al., 2017; Ray, 2016e). This may drastically lower down the cost of network processing and maintenance while supporting for green computing revolution (Ray, 2016f, 2016g, 2017a). Because, lower load on the active network would passive the effect of running the Hugh power hungry cloud data centers and network base stations in background (Chiang and Zhang, 2016; Canzian and Van Der Schaar, 2015; Ray and Agarwal, 2016; Ray, 2016h). The edge-IoT ecosystem can also pursue the geographically distributed aspects while allowing the mobility of end user. Surely, these two virtues give more expansion to its technological wings. It is also known that Content Delivery or Distribution Networks (CDNs) and Information-Centric Networking (ICN) work in similar fashion (Ray, 2015d; Elkhatib et al., 2017). But, they lack in interaction-free delivery services which has been given utmost importance in the edge-IoT ecosystem.
To this end, it is worth to mention that industrial standards and elements could also play a significant role when the edge-IoT ecosystem would be deployed in the reality (Farris et al., 2015; Elias et al., 2017). The reason is simple i.e. the use of popular and contemporary facilities which are currently accommodated with other computing services (Taleb et al., 2016; Kumar et al., 2013; Lin et al., 2015). For example, edge analytics server, NoSQL database engines, communication protocols, and data segments etc. Such industry elements have the power to fill the gap of the edge-IoT ecosystem to make it sharper and technologically visible (Xia et al., 2017; Dutta et al., 2016; Ray, 2017b). Till date, no such literature or document is available that has enlightened the fact of canvassing the industrial elements into the edge-IoT ecosystems. It is believed that upon such integration would provide immense power to the edge-IoT solutions to act efficiently on the undisclosed aspects. This survey presents an in-depth study and analysis of edge computing for IoT based scenario whereby implying the industry-protocols, standards, vendors, communication, data type, ecosystem and usage policies. Further, the presented survey is associated with a novel case study which makes this article unique in flavor and quality than the existing review or survey articles (Ray, 2016i; Hong et al., 2015; Jarschel et al., 2011; Kämäräinen et al., 2014; Fitzgerald et al., 2008).
The major contributions of this article are twofold. First, we present and discuss the taxonomical classification of the industrial edge-IoT computing. We review different factors of the taxonomy behind the edge-IoT. Second, we propose and evaluate a novel e-healthcare architecture that relies upon industrial elements of the edge-IoT ecosystem i.e. EH-IoT. Third, we discuss about the various pillars of the key requirements, functional capabilities, operational issues and architectural structure of the edge-IoT ecosystem to make it more strengthen in near future. The case study was specially demonstrated as a proof-of-concept to confirm that e-healthcare services can be harnessed from the EH-IoT. This indirectly advocates the efficiency of the industrial edge-IoT ecosystem. Many authors and researchers have effectively used and successfully incorporated edge computing paradigm to enhance robustness of their system. A quick search on the IEEE Xplore digital library with the keyword “edge computing”, “edge computing + e-health”, “edge computing + Edgent”, “edge computing + Apache Edgent”, “edge computing + Edgent + healthcare” gave a search result of 390 results out of which a handful of papers were found useful in context of applicability, parallelism and appropriateness. Some of these are discussed as follows: (Huang et al., 2014) validates the efficiency and resourcefulness of edge computing by providing extensive survey on edge systems and the also presents comparative study of cloud computing system. Their results show the edge system performs better than the cloud system. Ahlgren et al. (2012) developed a system which continuously monitors patient's heartrate with the help of heartrate sensor, this data is continuously analyzed by the KAA edge computing server, the analyzed data is sent to the user's smartphone. A design by Vallati et al. (Griffin et al., 2014) achieves a histrionic decrease in service latency and ensures the security of locality information. Sapienza et al. designed an edge computing-based smart urban-citizen application that can successfully perceive certain critical events e.g., man-made disasters (Ren et al., 2018). In Pan and McElhannon (2018) and Peterson et al. (2016), the researchers selected an edge computing-based architecture to address various network-related issues in vehicular technology. An efficient scheduling and adaptive offloading scheme is proposed that minimizes the computation complexities in the prescribed vehicular network. It is also known that the edge architectures might play a vital role in e-health applications which could save the lives of many patients. As edge computing guarantees a faster response and higher throughput, the decision-making process becomes faster and easier in the e-health applications. For instance, Ali and Ghazal (Ryden et al., 2014) proposed a real-time mobile detection service for heart attack by using the edge computing. The scheme showed lower service latency when it was combined with the geographical awareness that can rightly detect the patient's location. Lastly, Fl'avia Pisani et al. proposed their framework that is capable to execute cross platform code in the NodeMCU 1.0 (Habak et al., 2015).
The rest of this article is organized as follows: Section 2 discusses on the classifications of the edge-IoT taxonomy. Section 3 presents a novel EH-IoT architecture and test-bed is evaluated. Section 4 shows various parametric considerations required for sustainability for the edge-IoT ecosystems in terms of discussions. Lastly, Section 5 concludes the article.
Section snippets
State-of-the-art on the Edge-IoT taxonomy
This section presents an in-depth orientation on the state-of-the-art edge-IoT taxonomy. It covers eight classes of industrial components such as, (A) edge software and analytics, (B) edge-IoT ecosystem, (C) edge-IoT cloud platform, (D) edge-IoT key vendor, (E) edge-IoT data types, (F) edge-IoT open database systems, (G) Edge-IoT SoC platform and (H) Edge-IoT communication. More details on each of these components are prescribed later. Fig. 1 presents the taxonomy of industrial component of
Problem definition
Since Edge computing paradigm is a new concept, there is lack of availability of lightweight solution. Though many researches tend to implement this concept mostly by creating their own edge application which is specifically dependent to a particular hardware, they lack in open source aspect, ease of customization and applicability. Developers around the world is contributing for the development of open source software called Apache Edgent. Though, Edgent is still in its infancy stage, not much
Discussions and issues
This section enlightens edge-IoT architecture, functionality, operational issues, requirement, capability and selection criteria that are the most crucial factors for sustainability for the edge-IoT ecosystem in technological struggle.
Conclusion
Edge computing plays significant role in IoT applications where instant processing of data is required. This paper first reviews the existing industrial standards and solutions available for the edge-IoT. A novel edge-IoT architecture is prosed and discussed. Further, discussions are made to identify the requirement, capability and selection criteria of the edge-IoT ecosystem. Moreover, its architecture, functionality and operational issues are discussed. A novel case study while utilizing the
Acknowledgement
We are thankful to Mr. Arun Subba and Mr. Kiran Regmi for their support to conduct this research.
Partha Pratim Ray received the B.Tech. degree in computer science and engineering and the M.Tech. degree in electronics and communication engineering, with specialization in embedded systems, from the West Bengal University of Technology, Kolkata, India, in 2008 and 2011, respectively. He is currently a full-time Assistant Professor with the Department of Computer Applications, Sikkim University, Gangtok, India. His research interests include Internet of Things, Dew computing, and Pervasive
References (91)
- et al.
Fog and IoT: an overview of research opportunities
IEEE Internet Things J.
(2016) - et al.
Performance evaluation of remote display access for mobile cloud computing
Comput. Commun.
(December 2015) A survey of IoT cloud platforms
Future Comput. Inf. J.
(2016)Internet of things cloud enabled MISSENARD index measurement for indoor occupants
Measurement
(2016)- et al.
Green industrial internet of things architecture: an energy efficient perspective
IEEE Commun. Mag.
(2016) - et al.
A survey of information-centric networking
IEEE Commun. Mag.
(2012) - et al.
Real-time heart attack mobile detection service (RHAMDS): an IoT use case for software dependent networks
- et al.
Enabling cyber-physical communication in 5g cellular networks: challenges, spatial spectrum sensing, and cyber-security
IET Cyber-Phys. Syst.: Theor. Appl.
(2017) - et al.
Fog computing: a platform for Internet of Things and analytics
- et al.
Real-time stream mining: online knowledge extraction using classifier networks
IEEE Netw.
(2015)
Research on the predicting model of convenience store model based on digital twins
Clonecloud: elastic execution between mobile device and cloud
MAUI: making smartphones last longer with code offload
A survey of mobile cloud computing: architecture, applications, and approaches
Wireless Commun. Mobile Comput.
QoE-aware elasticity support in cloud-native 5G systems
Where's the bear?-automating wildlife image processing using iot and edge cloud systems
On using micro-clouds to deliver the fog
IEEE Internet Comput.
Federated edge-assisted mobile clouds for service provisioning in heterogeneous iot environments
AVE: autonomous vehicular edge computing framework with ACO-based scheduling
IEEE Trans. Veh. Technol.
Development and user evaluation of a virtual rehabilitation system for wobble board balance training
COMET: code offload by migrating execution transparently
Service oriented networking
Femto clouds: leveraging mobile devices to provide cloud service at the edge
Placing virtual machines to optimize cloud gaming experience
IEEE Trans. Cloud Comput.
GamingAnywhere: the first open source cloud gaming system
An evaluation of QoE in cloud gaming based on subjective tests
Towards pervasive and mobile gaming with distributed cloud infrastructure
A survey of computation offloading for mobile systems
Mob. Netw. Appl.
Leveraging edge analysis for internet of things based healthcare solutions
A survey on mobile edge computing: the communication perspective
IEEE Commun. Surv. Tutor.
Openflow: enabling innovation in campus networks
ACM SIGCOMM Comput. Commun. Rev.
Future edge cloud and edge computing for internet of things applications
IEEE Internet Things J.
Homecloud: an edge cloud framework and testbed for new application delivery
Central office re-architected as a data center (cord)
IEEE Commun. Mag.
Edge computing for the internet of things: a case study
IEEE Internet Things J.
Cited by (151)
Connectivity Standards Alliance Matter: State of the art and opportunities
2024, Internet of Things (Netherlands)Contemporary advances in multi-access edge computing: A survey of fundamentals, architecture, technologies, deployment cases, security, challenges, and directions
2023, Journal of Network and Computer ApplicationsSecurity paradigm for remote health monitoring edge devices in internet of things
2023, Journal of King Saud University - Computer and Information SciencesA review of IoT systems to enable independence for the elderly and disabled individuals
2023, Internet of Things (Netherlands)When machine learning meets Network Management and Orchestration in Edge-based networking paradigms
2023, Journal of Network and Computer ApplicationsTransformative synergy: SSEHCET—bridging mobile edge computing and AI for enhanced eHealth security and efficiency
2024, Journal of Cloud Computing
Partha Pratim Ray received the B.Tech. degree in computer science and engineering and the M.Tech. degree in electronics and communication engineering, with specialization in embedded systems, from the West Bengal University of Technology, Kolkata, India, in 2008 and 2011, respectively. He is currently a full-time Assistant Professor with the Department of Computer Applications, Sikkim University, Gangtok, India. His research interests include Internet of Things, Dew computing, and Pervasive bio-medical informatics. He received the VIRA Young Scientist Award and Bharat Vikas Award in 2017, for outstanding contribution in his field.
Dr. Dinesh Dash has received PhD from IIT Kharagpur, M.Tech from the University of Calcutta in 2013, 2004, respectively. Currently he is Assistant Professor in Computer Science and Engineering, National Institute of Technology, Patna, India. His area of interest includes Sensor Network, Mobile AdHoc Network, Algorithm, and Computational Geometry. He has served as the reviewer of IEEE Transaction on Mobile Computing and IEEE Access. He has received research grants from the Science & Engineering Research Board, DST Govt. of India.
Prof. Debashis De earned his M.Tech from the University of Calcutta in 2002 and his Ph.D (Engineering) from Jadavpur University in 2005. He is the Professor in the Department of Computer Science and Engineering of the West Bengal University of Technology, India, and Adjunct research fellow at the University of Western Australia, Australia.He is a senior member of the IEEE. Life Member of CSI and member of the International Union of Radio science. He worked as R&D engineer for Telektronics and programmer at Cognizant Technology Solutions. He was awarded the prestigious Boyscast Fellowship by the Department of Science and Technology, Government of India, to work at the Herriot-Watt University, Scotland, UK. He received the Endeavour Fellowship Award during 2008–2009 by DEST Australia to work at the University of Western Australia. He received the Young Scientist award both in 2005 at New Delhi and in 2011 at Istanbul, Turkey, from the International Union of Radio Science, Head Quarter, Belgium. His research interests include mobile cloud computing, Green mobile networks, and nanodevice designing for mobile applications. He has published in more than 200 peer-reviewed international journals in IEEE, IET, Elsevier, Springer, World Scientific, Wiley, IETE, Taylor Francis and ASP, seventy International conference papers, four researches monographs in springer, CRC, NOVA and ten text books published by Person education.