Congestion Aware Q-Learning (CAQL) in RPL Protocol – WSN based IoT Networks

In an Internet of Things (IoT) based real network scenario, the miscellaneous sensor applications are present which produce a heterogeneous traffic prototype. RPL (Routing Protocol for Low-power and Lossy Networks) is an attractive model for effective routing techniques in a heterogeneous traffic. The major objective functions (OFs) used in this prototype are objective function zero (OF0) and the Minimum Rank with Hysteresis Objective Function (MRHOF). But the routing pattern of standard OFs are not suitable for such network because it creates more congestion which leads to link failure and retransmission during communication. In our research, we observed the RPL protocol from base for heterogeneous traffic and proposed a novel routing protocol with a Reinforcement Learning Approach namely Congestion Aware Q-Learning (CAQL) in RPL Routing Protocol. CAQL technique is helpful for maintaining queue in optimal manner and load balancing between the nodes in the network to prevent the congestion, which lead to increase the performance of network in terms of Quality of Service (QoS). The implementation of the proposed model is done using the Network Simulator (NS2). The simulation of CAQL-RPL is done and the results are compared with the earlier protocol such as MRHOM, QWL-RPL. The performance evaluation shows that CAQL-RPL can provide better performance in a heterogeneous traffic prototype in terms of packet delivery ratio (PSR) and Packet loss ratio (PLR), Throughput, Energy Consumption and Network Delay.