Internet of Things (IoT) is a common Global platform to connect various heterogeneous devices over Internet which enable Machine-to-Machine (M2M) communication by embedding intelligent software in the objects to deploy high-quality and effective smart devices to offer different types of services. Congestion will be increased with the rate of increasing devices in the IoT networks. In congestion control mechanisms based on rate adjustment, dropping a packet will cause entire packet re-transmission due to packets are usually fragmented. In case of an alternative path approach, finding an optimal path with minimum hop count, low buffer occupancy, maximum link quality, low packet dropping rate and maximum through put is not easy to establish in every circumstance. In this paper, we propose an Adaptive Hybrid Congestion Control Protocol (AHCCP) for RPL based IoT sensor networks. In this protocol, initially the packets are sensed by an end IoT devices are categorized into various priorities depending on traffic type. Based on congestion status source will choose either rate based or alternative path for data transmission. This protocol will be implemented in NS3 and compare the results with game theory based congestion control protocol in terms of PDR, throughput, packet loss and energy efficiency.