Design of a Two-Tier Learning Model for Congestion Control in WSN Environment
D. Marygetsy, V. Remya, Nabeena Ameen, V. Anantha Krishna, M AnithaMary, M. Deivakani · 2022 International Conference on Edge Computing and Applications (ICECAA) · 2022
The Internet of Things (IoT) architecture and wireless networks provide more information about the covered region due to its high number of nodes. Due to the limitations in traditional data communication medium, several nodes are prone to transmission antenna collisions. Medium access control (MAC) approaches are commonly used in low-traffic networks, which are not susceptible to the noise generated from adjacent frequencies. The main aim of the proposed study is to develop solutions for preventing, detecting, and controlling network congestion. The proposed technique predicts the next step of the path in the congestion avoidance phase by distributing network traffic across optimal paths by using the Fuzzy-based decision-making method. During the congestion detection phase, a dynamic queue management mechanism is developed to swiftly identify congestion and prevent collisions. Back pressure based on queue quality was used throughout the packet scheduling process to lower the possibility of route connection from the pre-congested location. The major goal is to stabilise the network power consumption, reduce lost packet rates and enhance routing QoS.