Learning-Automata-Based Confident Information Coverage Barriers for Smart Ocean Internet of Things

• Published in: IEEE internet of things journal 2020-10, Vol.7 (10), p.9919-9929
• Main Authors: Deng, Xianjun, Jiang, Yalan, Yang, Laurence T., Yi, Lingzhi, Chen, Jiaoyan, Liu, Yong, Li, Xiangyang
• Format: Article
• Language: English
• Subjects: Algorithms; Barrier coverage; confident information coverage (CIC) model; Environmental monitoring; Internet of Things; Internet of Things (IoT); Learning automata; learning automata (LA); Machine learning; Marine environment; Nodes; Oceans; Optimization; Peer-to-peer computing; Sensors; smart ocean; Surveillance
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2020.2989696

As an emerging network paradigm, the Internet of Things (IoT) which consists of a significant number of multifunctional and heterogeneous IoT nodes has attracted dramatic attentions from both academia and industry. With the merits of intelligent capacity, desirable scalability, and high reliability, the IoT recently has been applied for smart ocean applications to provide protection for ocean environment monitoring and surveillance. Aiming to provide coverage service for ocean border environmental surveillance, this article studies the barrier coverage problem which investigates how to select a collection of IoT nodes to obtain an IoT node chain and build barrier paths to detect intruders and trespassers crossing the border region of interest. To overcome the disadvantages in the existing works on barrier coverage, we adopt a novel and widely adopted confident information coverage (CIC) model as the fundamental coverage model and formulate the CIC barrier path construction (CICBC) problem with the goals of maximizing the number of barrier paths and minimizing the amount of IoT nodes in each barrier path. We propose a distributed CIC barrier path (CICBP) construction approach based on learning automata (CBLA). The CBLA includes four crucial phases which are initialization phase, learning phase, monitoring phase, and repairing phase. Each IoT node equips a learning automaton. CBLA selects an optimal IoT node to construct the barrier path by learning. The simulation results show that the performance of the CBLA algorithm outperforms two peer algorithms in terms of the number of barrier paths and the average number of nodes in each barrier path.