Due to recent advancements in machine learning (ML) and cloud computing technologies, data has become far more useful and valuable. These technologies, in fact, make our lives easier and deliver smarter services. However, to enhance accuracy and efficiency of the service offerings, ML algorithms must learn from a large amount of training data. Therefore, beside the data itself algorithm design is the most crucial component of ML. Besides, cloud and edge computing are novel computational models that offer on-demand data storage, computational power, and infrastructure for ML algorithms. Therefore, data owners can outsource their data to the cloud servers for storage and training purposes. However, they will potentially lose control over their data and security of the data itself might be breached. Nevertheless, computation processing on un-encrypted sensitive data may jeopardize data confidentiality and might potentially expose users to various security threats including identity theft and fraud. In addition, for real-time services the offloading will create service quality issues in terms of longer latencies and users’ mobility.
For the last few years, our digital world of big data has seen an incredible increase from 1.2 zettabytes to almost 67 zettabytes of data. Similarly, emerging technologies and diversity of user devices which form part of the internet of things (IoT), multi-access edge clouds (MECs), and modular applications are beginning to also become dominant. Furthermore, the rapid increase in real-time applications such as, online games, movies, educational services, will create huge network traffic and processing requirements that will affect service quality and computing economics at scale. Therefore, the amount of data and computational processing issues are the key concerns in ML and edge-cloud computing. In fact, new computational and learning strategies should be developed to fulfil the demand for intelligent computation in edge-cloud computing using artificial intelligence (AI) and ML methods. Although, AI techniques have been widely used for resource management in the cloud, but relatively unexplored in the edge-cloud continuum. Similarly, traditional cryptographic algorithms can secure data secrecy, however these methods cannot be used to compute over the encrypted data. As a result, these approaches are incompatible with ML, AI, and edge-cloud computing. Topics of interest include (not limited to):
- Resource allocation, consolidation, and orchestration to optimize edge-cloud infrastructure
- Energy, performance, and cost-efficient edge, cloud service offerings
- AI and ML algorithms for intelligent computation over the edge-cloud infrastructure
- Partitioning applications across edge and cloud environments (using privacy constraints)
- Data and network management to facilitate fog, edge, and cloud integration
- Big data, IoT at edges, and AI based optimization techniques
- Resource management to enable intelligent edge-cloud computing
- Secure data transmission, offloading, and privacy-preserving methods
- Techniques for management of data storage at the edge-cloud
- Using edge-cloud to support healthcare, intelligent agriculture, smart cities etc.
- Machine learning based frameworks for edge and cloud computing, and
- Applications for big data intelligence leveraging edge-cloud and fog computing.
This SI aims to present state-of-the-art, research challenges, solutions, and applications for intelligence in edge-cloud computing. It also aims to cover various aspects of resource management and ML based framework that supports intelligence in edge-cloud continuum. The outcome will be a collection of articles that propose techniques for digitizing various services in the domain of edge-cloud computing, machine learning, and AI.
Tentative Timeline
Deadline for submissions: 31st January 2023
1st Review: 30th November 2022
2nd Review: 30th January 2023
3rd Review: 28th February 2023
Final Decision: 30th March 2023
Guest Editors
Dr. Muhammad Zakarya (Ph.D., SMIEEE)
Assistant Professor, Program Director, iFuture: a leading research group, Head of the Cloud computing, Big data, and Intelligence (CBI) Laboratory, Abdul Wali Khan University, Pakistan
https://awkum.edu.pk/
mohd.zakarya@awkum.edu.pk
Prof. Jinguang Han (Ph.D., postdoc)
Professor, School of Cyber Science and Engineering, Southeast University, China
https://www.seu.edu.cn/english/
jghan@seu.edu.cn
Dr. Santosh Tirunagari (Ph.D., postdoc)
Assistant Professor, Middlesex University, UK
https://www.mdx.ac.uk/
s.tirunagari@mdx.ac.uk
Dr. Atta Ur Rehman Khan (Ph.D., SMIEEE)
Associate Professor, College of Engineering and Information Technology, Ajman University, UAE
https://www.ajman.ac.ae/en
a.khan@ajman.ac.ae