The Internet of Things (IoT) connects various areas by connecting millions of devices to meet a wide range of human needs. A huge amount of data has to be transmitted for these services. In this transmission, Internet of Things (IoT) networks do not provide any special priority for emergency data packets while routing. Using conventional QoS processes, these data packets flow through routers. This transmission does not guarantee that an emergency data packet traveling through a congested IoT network will reach the control room in time. One of the major challenges in packet scheduling is the unpredictable behavior of traffic classes, which change dynamically. For this reason, to overcome prioritization problems in IoT networks, innovative packet prioritization techniques, such as a queue management approach, need to be developed. To provide the required transmission priority for emergency data, this paper proposes an AI packet priority queuing model (PPQM) based on P2 queue-based emergency data packet classification with a prioritization algorithm. In this paper, LSTM is used to classify the emergency data packet, and the Deep Q Network (DQN) algorithm is proposed to make scheduling decisions for communication. Simulation results confirmed that the machine learning modules achieved 91.5% accuracy while identifying the emergency data and assigning them the expected priority.