In this paper, a high-availability data collection scheme based on multiple autonomous underwater vehicles (AUVs) (HAMA) is proposed to improve the performance of the sensor network and guarantee the high availability of the data collection service. Multi-AUVs move in the network and their trajectory is pre-defined. Data from the nodes near the trajectory of an AUV are sent to the AUV directly and the other nodes transmit their data to the nodes that are closer to the trajectory. A malfunction discovery and repair mechanisms are applied to ensure that the network operates appropriately when an AUV fails to communicate with the nodes while collecting data. Compared with existing methods, the HAMA method increases the packet delivery ratio and the network lifetime

In this paper, an autonomous underwater vehicle (AUV) location prediction-based data collection scheme (ALP) has been proposed, to overcome high and unbalanced energy consumption for underwater wireless sensor networks (UWSNs). In our scheme, an AUV travels around the network, follows a predefined trajectory, and collects data from the sensor nodes. The nodes near the trajectory send their data to the AUV directly while the others send data to their neighbors that are closer to the trajectory. To overcome the “hot region” problem that means the nodes near the trajectory of the AUV to consume energy faster and die early, a trajectory adjustment mechanism is applied. A mathematical model is proposed to adjusted the trajectory periodically. To guarantee an efficient communication between the nodes and the AUV, a reliable time mechanism is proposed. In this mechanism, only the nodes with a sufficient amount of time are capable to communicate and send their data to the AUV directly. The analyses and simulation results validate that our proposed ALP prolongs the network lifetime and has a higher packet delivery ratio than the existing protocols.

Challenges in operating system courses, such as inefficient evaluation of practical skills, delays in debugging, and a lack of personalized instruction, were addressed by an AI-driven platform developed in this study. The platform enables precise teaching and efficient debugging through real-time compliance verification, intelligent error diagnosis, and dynamic generation of personalized tasks. By utilizing full-process behavioral data for multidimensional competency assessment, the platform was deployed in over 20 classes. It effectively reduced instructors’ workload while enhancing students’ theoretical knowledge and practical understanding, promoting efficient, personalized, and intelligent transformation in operating system training.