Coap Based Congestion Control Mechanisms For Internet of Things

Abstract

Internet of Things (IoT) is a network where physical objects with Internet connectivity can interact and exchange information with other connected objects. IoT devices are con- strained in terms of power and memory, and have limited communication capabilities. A large number of IoT devices are characterized by small memory and low processing speeds that lead to congestion in the network when many such devices try to communicate with each other. Typically, TCP is responsible for end-to-end congestion control and reliability. However, in a constrained network, this can cause performance problems because most of the communication happens over wireless links that are known to be challenging for TCP. Consequently, the responsibility for controlling congestion is entrusted to the application layer protocols. Among several IoT application layer protocols, only the Constrained Application Protocol (CoAP) has a built-in congestion control mechanism and has been standardized by the IETF. CoAP is a lightweight messaging protocol which is widely used by various IoT appli- cations in low power and lossy wireless networks. CoAP provides reliability and minimal congestion control via a fixed Retransmission TimeOut (RTO) and Binary Exponential Backoff(BEB). It does not maintain end-to-end connection information and therefore, cannot adapt RTO based on the network conditions. CoAP Simple Congestion Con- trol/Advanced (CoCoA) is an improved congestion control mechanism for CoAP which adapts RTO based on the network conditions. Congestion control mechanisms for CoAP can be classified into non-RTT and RTT based mechanisms. Non-RTT based mechanisms (e.g., CoAP) are advantageous for applications that require a low memory footprint and do not maintain the state information. RTT based mechanisms (e.g., CoCoA) are useful for applications wherein memory footprint is not a major concern and maintaining the state information is desirable. Recently, it has been shown that the congestion control techniques in CoAP and Co- CoA require further optimizations. In this work, we thoroughly evaluate the congestion control mechanisms in CoAP and CoCoA, and subsequently propose one non-RTT and i three RTT based optimizations: (i) Geometric Sequence Technique for Effective RTO Estimation in CoAP (GST-CoAP), (ii) Effective RTO estimation using Eifel Retransmis- sion Timer in CoAP (CoAP-Eifel), (iii) Geometric Series based effective RTO estimation Technique for CoCoA (GSRTC), and (iv) CoCoA++. GST-CoAP is a non-RTT based mechanism to improve RTO estimation in CoAP by using geometric sequence. CoAP-Eifel, GSRTC and CoCoA++ are RTT based mech- anisms. CoAP-Eifel integrates the Eifel Retransmission Timer with CoAP to improve RTO estimations and control congestion. GSRTC is a simple enhancement which adapts the weight used in the Strong RTO estimator instead of using the fixed weight (0.5). CoCoA++ uses delay gradients to get a better measure of network congestion, and im- plement a probabilistic backoffto deal with congestion. A common characteristic in the proposed mechanisms is that they are easy to deploy and do not add user configurable parameters. All the four proposed mechanisms are extensively evaluated in a wide variety of network environments using the Cooja simulator and a real testbed at FIT/IoT-LAB. The performance metrics used for evaluation are Flow Completion Times (FCT), number of retransmissions, throughput, RTO, delay and packet sending rate. We observe that the proposed mechanisms offer a notable improvement in network performance compared to the existing mechanisms.