The Internet of Things has been heralded as the most important technical development in recent years due to its low-cost and low-power sensor technology. The IoT makes everyday tasks easier, including smart transportation, smart housing, and smart healthcare, which uses Bluetooth and other low-power wireless technologies in low power lossy network (LLN). The LLN has low throughput and substantial delay due to its traffic patterns, to communicate with resource-constrained sensor devices. To fix this issue RPL is used to solve complicated problems. Routing Protocol for Low-Power and Lossy Networks (RPL) is still working through issues with energy use, scalability, security, and dependability, to name just a few. However, managing all of these challenges in the RPL-IoT network is essential since the RPL contains heterogeneous traffic. In this paper, a novel approach on Deep Sequential Neural Network (DSNN) based Authenticated Routing Enforced RPL (DARE-RPL) is introduced to overcome complex issues in routing and secure transmission of packet. The border router for the LLN network additionally connects the roots to the internet. It has been suggested to use a secure RPL with a congestion avoidance approach to lessen traffic and offer a secure environment. DSNN is the congestion detection method used to avoid congestion in the network during the transmission. The MATLAB simulator is use to evaluate the DARE-RPL technique. The load balancing capacities of the RPI-IOT and MADM current systems are lower for the same set of malicious nodes, at 69% and 58%, respectively, compared to the 93% load balancing capacity achieved by our proposed DARE-RPL.