Congestion Control Scheme for Link Quality Improvement Using Bio-Inspired Fuzzy Inference System–Based CoAP in IoT

Abstract

The Internet of Things (IoT) represents the Internet's future by including devices that can connect with one another. IoT devices in IoT networks collect data from their surroundings and send it in bursts to a server in a remote location. The quantity of applications using computer networks causes resource competition, which in turn causes congestion. The most difficult task in enhancing network quality of service (QoS) is IoT congestion control. For IoT items with modest resource requirements, various protocols have arisen. Different shortages and issues affect basic congestion control, which increases bandwidth use, data loss, and delay. Thus, to solve this issue, in this paper, we propose a bio-inspired fuzzy inference system (FIS)–based constrained application protocol (BIFIS-CoAP) for avoiding congestion over network. In order to determine the level of congestion, we use the bottleneck bandwidth gradient (BG-gradient) and round-trip time gradient (RT-gradient) as inputs for BIFIS-CoAP. By using this indication, BIFIS-CoAP may anticipate impending congestion and alter the sending rate to prevent it. In order to achieve high performance for burst data transfer, BIFIS-CoAP constantly checks for bandwidth that is available and uses the congestion degree for updating the variable retransmission time out (RTO) for retransmissions. Using the adaptive Tasmanian devil optimization method (ATDO), input parameters like the RT-gradient and BG-gradient are modified to enhance the performance of the FIS. Numerous simulation studies have shown that BIFIS-CoAP is feasible, and the simulation results show that it performs better than basic CoAP and fuzzy CoAP in terms of delay, throughput, delivery ratio, and retransmissions.