An Adaptive Congestion Control Algorithm with CoAP for the Internet of Thing

CoAP is a dominant application protocol designed for the constrained devices in IoT (the Internet of Things) which is going to interconnect every physical object in our daily life together to make an intelligent world. To keep the communication between devices in IoT from congestion, CoAP originally provides a simple congestion control mechanism. Since this mechanism can hardly meet the requirements of many IoT applications, an extension called CoCoA is being standardized by IETF(the Internet Engineering Task Force) to optimize CoAP. However, the thresholds underpinned the VBF(variable backoff factor) and RTO(retransmission timeout) aging mechanism in traditional CoCoA are constants and cannot be changed under most conditions which limit the performance of CoCoA; and further the existed estimators of CoCoA ambiguity have the potential risk to cause some unexpected or unpredictable problems. Hence, in this paper, we propose a novel threshold adjustable scheme for the VBF and RTO aging mechanism and a new encapsulation method for message retransmission to solve the above problems in the traditional CoCoA. Our evaluation results show that our modified version CoCoA greatly helps to solve the problem of congestion in IoT which outperforms over the existed CoCoA in terms of throughput and network adaptability.