Special Issue on Deep learning for Intelligent IoT: Opportunities, Challenges and Solutions
摘要截稿:
全文截稿: 2019-09-30
影响因子: 2.816
期刊难度:
CCF分类: C类
中科院JCR分区:
• 大类 : 计算机科学 - 3区
• 小类 : 计算机:信息系统 - 3区
• 小类 : 工程:电子与电气 - 3区
• 小类 : 电信学 - 3区
Overview
Future-generation wireless networks are required to support ultra-reliable and low-latency communications. In past few years, artificial intelligence (AI) flourished by advancements in machine learning (ML) more specifically in deep learning (DL) and reinforcement learning (RL) showing their value in a wide range of applications, where classification or regression problems play a pivotal role. One of the key-applications of future-generation wireless networks is the management of the Internet of things (IoT) infrastructures. The complex, heterogeneous and distributed nature of this type of systems leads many researchers and practitioners to explore the application of DL techniques to make IoT intelligent, reliable and highly performing. Forbes predicted that IoT industry will grow to $457 billion by 2020. Smart cities, smart grid, health and industrial IoT (IIoT) are the key contributors to this massive adaptation and growth. Specifically, the revolution known as Industry 4.0 takes advantage from the developments in the IoT field enormously.
IoT systems consist of a large number of distributed heterogeneous devices that generates consistent amount of data. Cleary, wireless solutions have a primary importance in this context as witnessed by the massive development of industrial (often low-cost) components that have been helping the actual implementation of this technology. Data harvested by the devices is progressively increasing in size and heterogeneity. Moreover, IoT network is diverse and complex in nature. Algorithms devised for networks with certain structured characteristics are not typically efficient once applied to IoT systems. Finally, IoT devices also possess limited computational, memory and energy resources so they heavily rely on edge and core networks for data handling, processing and analysis. As a consequence, DL-based edge and core devices are crucial to develop an intelligent and efficient resource management and network management and to improve the overall system performance. Indeed, the quantitative evaluation and optimization of IoT system is fundamental area of research: The wireless traffic is growing exponentially with the continuous increase of the number of IoT devices that generate massive short-packet transmissions over the wireless downlink/uplink. For these reasons, communication protocols also need to be adaptive to such a dynamic IoT environment in real time. With respect to ordinary wired and wireless networks, IoT networks usually require to be optimized in terms of their energy consumption and load-balancing strategies have to be devised to run the IoT devices for many years in a reliable way.
From the functional point of view, the heterogeneity of the IoT devices also raises interoperability concerns. The anomaly detection and defense is also one of the huge problems for the industry: protecting the infrastructure from malicious network attacks, unauthorized access and safeguarding the user privacy are fundamental, but at the same time challenging requirements. For example, the anomaly detection systems require to adapt themselves to unpredictable events in order to cater the largest class of cyber threats. Intelligent IoT requires to cope all the aforementioned enormous challenges. DL promises to solve such a diverse range of problems without human intervention.
This special issue aims to bring together the academia and industrial researchers to explore the opportunities of DL for IoT, study its impact on the solution of the solution of the aforementioned challenges and propose viable solutions. We solicit papers covering various topics of interest that include, but not limited to the following.
DL-based architecture and technologies for intelligent IoT
DL-based services for smart cities, grids and health for intelligent IoT
DL-based multimedia services for intelligent IoT
DL-based big data mining and analytics for intelligent IoT
DL-based applications for intelligent IoT
DL-based transport protocols for intelligent IoT
DL-based routing protocols for intelligent IoT
DL-based MAC protocols for intelligent IoT
DL-based radio duty cycling (RDC) for intelligent IoT