Special Issue on Sustainable Computing, Communication and Networks for Large-scale Artificial Intelligence

 Introduction:

As the number of parameters in artificial intelligence (AI) models continues to increase, the demand for computing resources and training time grows. This necessitates the use of high-performance computing clusters and distributed training techniques. Moreover, real-time data transmission becomes essential for device-edge-cloud collaboration. To meet these requirements and facilitate data transmission and exchange, high-speed and reliable communication networks like 5G and Wi-Fi play a crucial role.

In distributed applications, effective connectivity between different computing nodes via the network is necessary for model training and deployment. This calls for the use of high-speed and reliable network connections to ensure the stability and reliability of data transmission. Furthermore, when deploying the model, optimizing network topology and implementing load balancing techniques are essential considerations to ensure efficient and reliable model operation. In conclusion, the efficient, stable, and reliable development of large-scale AI models requires support from computing, communication, and networks.

In response to the above-mentioned issues, researchers are addressing the challenges of large-scale AI computing, communication, and networking through sustainable solutions. Hardware optimization includes developing low-power specialized processors and neuromorphic chips, while software optimization involves energy-efficient algorithms and code tailored to specific hardware. At the same time, there are some other good solutions on the above-mentioned issues as follows. Advanced networking protocols like software-defined networking and network function virtualization enable resource allocation and data transmission optimization. Wireless technologies such as Wi-Fi and 5G support energy-efficient communication between edge devices and AI models. Edge computing reduces data transmission and energy consumption. Joint optimization of computing, communication, and networking involves task coordination, hardware-software co-design, and efficient data distribution. Renewable energy sources like solar and wind power can further reduce the carbon footprint of large-scale AI applications.

This special issue aims to serve as a platform for showcasing advanced theories and applications in the field.

Topics covered include, but are not limited to:

• Energy-efficient computing architectures and hardware for large-scale AI models
• Novel techniques for reducing the computational complexity of large-scale AI models
• Communication networks and protocols for big data transmission and model training
• Integration of renewable energy sources in large-scale AI communication infrastructure
• Sustainable communication and networking protocols for large-scale AI systems
• Distributed AI algorithms for sustainable computing and communication
• Green computing and networking solutions for large-scale AI
• Sustainable data centers and cloud infrastructures for large-scale AI systems
• Energy-efficient hardware and software design for large-scale AI systems
• Energy harvesting and power management for sustainable and large-scale AI systems
• Network protocol decision-making and management for large-scale AI systems
• Device-edge-cloud network communication optimization for large-scale AI systems
• Network virtualization and network architecture design for large-scale AI systems
• High performance network communication architecture for large-scale AI systems
• Standardization of network communication for sustainable and large-scale AI systems
• Openness and scalability of network protocols for sustainable and large-scale AI systems
• Case studies and real-world applications of large-scale AI models

Important dates:

• Submission deadline: December 30, 2023
• First notification: March 15, 2024
• Revised version deadline: May 15, 2024
• Final notification: June 30, 2024
• Publication date: 2024

Submission instructions:

Please read the Guide for Authors before submitting. All articles should be submitted online via the editorial management system; please select article type VSI: SCCN4LAI.

Guest Editors:

• Prof. Wei Wei (lead Guest Editor), Xi’an University of Technology, China. Email: weiwei@xaut.edu.cn
• Prof. Shahid Mumtaz, Nottingham Trent University, UK. Email: dr.shahid.mumtaz@ieee.org
• Prof. Muhammad Khurram Khan, King Saud University, Saudi Arabia. Email: mkhurram@ksu.edu.sa