Call for Articles

The Internet of things (IoT) is an interconnected system of computing devices, machinery, and digital machines that digitize the real world. The IoT has already affected people's lives, including transportation, housing, food, clothing, health, and remote monitoring. Many home appliances can be controlled through mobile phones and voice. Many applications allow users to improve their quality of life, and even enable the elderly and the disabled to live more conveniently. MGI's report shows that starting from 2025, the Internet of Things will create an output value of 3.9 trillion to 11.1 trillion US dollars in nine environments, including factories, retail, and cities, and the number of IoT devices is expected to grow to 754 100 million, which is equivalent to adding 127 IoT devices every second in the world starting in 2020. The operation of IoT systems can be summarized by the following three phases: the deployment of sensors to collect data, the conversion of collected data into useful information being able to be stored and accessed, the transformation of information to domain knowledge which will be utilized by the IoT system controller for users’ feedback or the system reactions. An IoT system becomes an intelligent IoT system if all tasks involving the three phases of IoT operations can be automated.

Data science (DS) is a multidisciplinary approach to discovering, extracting, and presenting insights in data by focusing on data collection, data store and access, data analysis, and data communication techniques. Data science includes descriptive, diagnostic, predictive, and prescriptive capabilities. This means that through data science, administrators can use data to figure out what happened, why it happened, what happened, and what they should do with expected outcomes. Since the automation of an intelligent IoT system requires all tasks of DS, DS will be the most proper candidate technology ready to solve those issues faced by intelligent IoT systems. To collect data for IoT applications features, how to design sensor deployment and their connections via communication networks is the first main problem for intelligent IoT. The next step is how to apply machine learning (ML) and artificial intelligence (AI) algorithms to analyze and interpret insights concerning collected intelligent IoT data. Finally, it is also very crucial to communicate analysis results effectively to users of intelligent IoT devices.

From the viewpoint of DS, we believe that the following categories of problems should benefit from DS related technologies in developing future intelligent IoT systems. The first problem is how to deal with intelligent IoT Big Data. The amount of generated data in each application unit of an intelligent IoT system is at least the scale of Terabytes (TB). The collection, exchange, storage, and access of such a huge amount of data at an intelligent IoT device is an exceedingly challenging task, as the computational and communication resources in intelligent IoT devices are extremely limited. Deep learning is a breakthrough technology in ML/AI. Deep-learning (DL) applications on IoT devices often have an extremely strict real-time requirement. For example, security camera–based object-recognition tasks usually require a detection latency of less than 400 ms to capture and respond to target events—for example, abnormal targets (identified by DL technology) appearing inside a building—in a short response time. Current IoT devices often offload intelligence computation to the cloud. However, consistent and reliable wireless communication links, which are only available at limited locations with high cost, become one of the main difficulties for these intelligent IoT devices to fulfill real-time requirements. Hence, the second category of problems about intelligent IoT is to have advanced ML/AI algorithms which can perform data analysis with input data impacted by unreliable communication links. However, enabling ML/AI capabilities on the intelligent IoT device side is not an easy assignment. The main properties of intelligent IoT devices are small memory size, low power and distributed. The third category of problems are to design new ML/AI algorithms that can be implemented at IoT devices in a distributed manner under small memory size and low power constraints. Finally, security, trust and privacy of intelligent IoT users are always main considerations for any new technologies. With the huge number of intelligent IoT connected devices, how to apply DS to enhance access control systems, trust management, and secure data sharing with privacy considerations over intelligent IoT systems is a challenging problem.

This Special Issue (SI) will highlight the multidisciplinary approach to discovering, extracting, and presenting insights in data by focusing on data collection, data store and access, data analysis, and data communication techniques. Data science includes descriptive, diagnostic, predictive, and prescriptive capabilities. This means that through data science, administrators can use data to figure out what happened, why it happened, what happened, and what they should do with expected outcomes.

Topics of interest include, but are not limited to:

• Data science driven intelligent IoT system architecture design.
• Intelligent IoT Big Data collection, storage and access.
• Distributed data analysis.
• Design new ML/AI algorithms to train at intelligent IoT devices under insufficient and contaminated data.
• Implement advanced ML/AI algorithms in a distributed manner under small memory size and low power constraints.
• Intelligent IoT data integrity, confidentiality and availability problems identifications and countermeasures.
• Threat models and counterattack strategies for intelligent IoT.
• Distributed and resource-saving intrusion detection systems for intelligent IoT.
• Reliability and reputation model for the trust level of intelligent IoT devices.
• Performance and scalability analysis in data science driven intelligent IoT.
• Human-Computer Interaction (data visualization) in application development for data science driven intelligent IoT.
• Application of intelligent IoT to social IoT.
• Application of intelligent IoT to smart city.
• Application of intelligent IoT to industrial IoT.
• Leveraging the data-science driven intelligent IoT to other fields applications: energy, smart grid, logistics, transportation, supply chain, monetization, e-business, notarization, e-government, e-health, e-commerce, insurance, finance, fintech, e-learning, crowdsourcing, and crowd sensing applications.

Submission Guidelines

Manuscripts should conform to the IEEE Internet of Things Magazine standard format as indicated in the Information for Authors section of the Article Submission Guidelines. All manuscripts to be considered for publication must be submitted by the deadline through the magazine’s Manuscript Central site. Select “June 2023/ Data Science Driven Intelligent IoT” from the drop-down menu of Topic/Series titles.

Important Dates

Manuscript Submission Deadline: 14 February 2023 (Extended Deadline)
First Decision Date: 1 February 2023
Final Revisions Due: 15 March 2023
Final Decision Date: 1 April 2023
Final Manuscript Due: 15 April 2023
Guest Editorial/Column: 22 April 2023
Expected Publication Date: June 2023

Guest Editors

Pin-Han Ho (Lead Guest Editor)
University of Waterloo, Canada

Shih Yu Chang (Corresponding Guest Editor)
San Jose State University, USA

James She
Hamad Bin Khalifa University, Qatar

Yuren You
Huawei Canada Technologies, Canada

Min Chen
Huazhong University of Science and Technology, China

János Tapolcai
Budapest University of Technology and Economics, Hungary