[Zurück]


Zeitschriftenartikel:

P. Donta, T. Amgoth, C. Annavarapu:
"Delay-aware data fusion in duty-cycled wireless sensor networks: A Q-learning approach";
Sustainable Computing-Informatics & Systems, Volume 33 (2022), S. 100642: 1 - 100642: 15.



Kurzfassung englisch:
In wireless sensor networks (WSNs), the sensor nodes (SNs) are deployed to acquire the data from the area of interest and transmit it to the sink via multi-hop communications. Due to computation, buffer, and energy constraints, the SNs need efficient routing to forward the data in time to sink with limited energy drain, and it is a challenging task. It is more difficult in duty-cycled WSNs because the SNs are active for a limited time and inactive in the remaining time to minimize the energy drain. In this context, we propose a delay-aware data fusion (DADF) approach to the trade-off between the delay and energy while performs the data fusion. Initially, the DADF performing the data fusion operation to avoid duplicating and inconsistent data at each SNs using a simple statistical approach during its active state. Afterward, the sink uses reinforcement learning to identify the best forwarding node of each SN for data communications with minimum delay and energy in duty-cycled WSNs. Each forwarding node operates the data fusion once if it acquires the data from its child nodes. The proposed method using various performance metrics such as network lifetime, throughput, energy consumption, buffer utilization, and end-to-end delay are compared with recent and relevant existing techniques, and our methods outperform them in varying dynamic conditions.

Schlagworte:
Data fusion, Delay-aware, Duty-cycle, Internet of Things, Network lifetime, Q-learning, Throughput, Wireless sensor networks


"Offizielle" elektronische Version der Publikation (entsprechend ihrem Digital Object Identifier - DOI)
http://dx.doi.org/10.1016/j.suscom.2021.100642


Erstellt aus der Publikationsdatenbank der Technischen Universitšt Wien.