The Influence of Headphones on the Localization of External Loudspeaker Sources

Wireless sensor networks have become incredibly popular due to the Internet of Things' (IoT) rapid development.IoT routing is the basis for the efficient operation of the perception-layer network.As a popular type of machine learning, reinforcement learning techniques have gained significant attention due to their successful application in the field of network communication.In the traditional Routing Protocol for lowpower and Lossy Networks (RPL) protocol, to solve the fairness of control message transmission between IoT terminals, a fair broadcast suppression mechanism, or Drizzle algorithm, is usually used, but the Drizzle algorithm cannot allocate priority.Moreover, the Drizzle algorithm keeps changing its redundant constant k value but never converges to the optimal value of k.To address this problem, this paper uses a combination based on reinforcement learning (RL) and trickle timer.This paper proposes an RL Intelligent Adaptive Trickle-Timer Algorithm (RLATT) for routing optimization of the IoT awareness layer.RLATT has triple-optimized the trickle timer algorithm.To verify the algorithm's effectiveness, the simulation is carried out on Contiki operating system and compared with the standard trickling timer and Drizzle algorithm.Experiments show that the proposed algorithm performs better in terms of packet delivery ratio (PDR), power consumption, network convergence time, and total control cost ratio.