Future Internet of Things (IoT) communication trends toward heterogeneous services and diverse quality-of-service requirements pose fundamental challenges for network management strategies. In particular, multiobjective optimization (MOO) is necessary in resolving the competition among different nodes sharing limited wireless network resources. A unified coordination mechanism is essential such that individual nodes conduct the opportunistic maximization of heterogeneous local objectives for efficient distributed resource allocation. To such a problem, this article proposes an artificial intelligence (AI)-based framework, which is termed as MOO strategy for AI-aided IoT communications (MOSAIC). This framework enables to tackle numerous MOO tasks in IoT network management with simple reconfiguration of learning rules. In this strategy, a component unit associated with an individual network node includes a pair of deep neural networks (DNNs) to learn optimal local functions responsible for calculation and distributed coordination, respectively. The resultant AI module swarm called DNN tiles realizes the node cooperation that collectively seeks distributed MOO calculation rules. The advantage of MOSAIC is characterized by Pareto tradeoffs among conflicting performance metrics in diverse wireless networking configurations subject to severe interference and distinct criteria for multiple targets.
Bibliographical noteFunding Information:
This work was supported in part by the NRF Grant Funded by the Korea Government Ministry of Science and ICT (MSIT) under Grant 2021R1I1A3054575; in part by the Institute of Information and Communications Technology Planning and Evaluation (IITP) Grant Funded by MSIT (Intelligent 6G Wireless Access System and Research on LEO Inter- Satellite Links) under Grant 2021-0-00467 and Grant 2021-0-00260; in part by the Korea University Grant; and in part by the National Research Foundation, Singapore and Infocomm Media Development Authority under its Future Communications Research and Development Programme.
© 2022 IEEE.
- Deep learning (DL)
- distributed network management
- multiobjective optimization
- primala-dual training
ASJC Scopus subject areas
- Signal Processing
- Information Systems
- Hardware and Architecture
- Computer Science Applications
- Computer Networks and Communications