Adaptive Resource Observation for Congestion Alleviation using Constrained Application Protocol

This paper presents an adaptive resource observation (ARO) for congestion alleviation using constrained application protocol (CoAP), which prevents buffer overflow of the client by adjusting observing period of the associated servers. The operation of ARO consists of two main phases; 1) buffer overflow estimation, 2) observing period adaptation. In the former, the client estimates whether buffer overflow will occur by comparing its service rate with packet arrival rate, then it determines the new observing period that can prevent buffer overflow of the client. The latter is used to adjust the observing period of servers considering the predefined the minimum and maximum queue threshold. ARO can significantly reduce the number of dropped packets caused by buffer overflow. The simulation results show that ARO achieves a higher network performance than legacy CoAP.