MDPI IoT Journal (ISSN 2624-831X)

Special Issue "Edge Computing Optimization Using Artificial Intelligence Methods"

The growing importance of the Internet of Things (IoT) and the ubiquitous high capacity provided by 5G technologies have brought the specter of massive quantities of data being generated and/or consumed by sensors, actuators, and smart devices. Such massive amounts of data require considerable processing power, which is available in the cloud. However, cloud-based computation and data delivery models do not allow the stringent quality of service (QoS) guarantees to be efficiently harnessed. The latter is due to the number of hops of wired networks between the data endpoints and the cloud, which leads to a significant increase in latency, which may dramatically affect real-time control and other critical systems. Moreover, forwarding all the data generated by such devices directly to the cloud may devour the network bandwidth, leading to congestion. Therefore, it is necessary that critical processing to be hosted closer to the endpoint devices, i.e., closer to the sources and sinks of the data so that data can be processed and filtered out by the time it reaches the cloud. This can be achieved through Edge Computing (EC).

Efficient, scalable, and QoS-aware placement of IoT data processing jobs in EC resources is a complex optimization problem and, currently, an active research topic. As new jobs are created, they have to be assigned computational resources dynamically, matching job requirements with the cost, reliability, location (and mobility), besides the current availability of the resources. Less critical or demanding communication jobs may be offloaded to the cloud. The use of Artificial Intelligence (AI) methods to jointly tackle the problem of job placement optimization, including jobs belonging to AI-based data analytics software, constitute currently active research topics addressed by this Special Issue.

For this Special Issue, original scientific articles are welcome on the following as well as closely related topics:
- AI-based algorithms to optimize job placement in EC
- AI software architectures favoring distributed computing job placement in EC resources (e.g., Distributed Deep Neural Network architectures)
- AI-based mechanisms supporting open EC markets leveraging the participation of third-party computing resources opportunistically (e.g., parked autonomous vehicles)
- AI-based methods to optimize mobile EC resources' placement (e.g., EC capable drones)


Guest Editors:
- Prof. Dr. António M.R.C. Grilo
- Prof. Dr. Paulo Rogerio Pereira
- Prof. Dr. Naércio Magaia