AdCoCoA- Adaptive Congestion Control Algorithm for CoAP

Abstract

Internet of things (IoT) is a system of interconnected computing devices, mechanical, digital devices, objects with unique identifiers, and able to communicate over a network without any intervention. Congestion control is one of the critical challenges in IoT networks. The reasons which cause congestion in the IoT networks are unnecessary retransmissions, packet collisions, lossy links, and higher packet error rates. The Internet Engineering Task Force (IETF) has standardized CoCoA+ protocol for CoAP. CoCoA+ derived from LinuxRTO, performs better than basic CoAP in many scenarios. But performs significantly worse in light traffic and bursty traffic environment. In this paper, we present an adaptive congestion control protocol for CoAP derived from CoCoA+, named Adaptive Congestion Control Algorithm (AdCoCoA). The proposed algorithm estimates RTO using dynamic factors instead of fixed values. This dynamic adaptation helps to improve CoAP performance and reduce retransmissions. The results show AdCoCoA has lower RTO values, retransmission index (11.6%); higher goodput (71.3%), packet delivery ratio (36.4%), and transmission rate (55.6%) compare to CoCoA+.