Home Page Diaries**Hobbies Learning Notes**Academics Leave Your Cmts.**More Info.<aside> 💡 This article contains machine-translated parts, please check carefully.
</aside>
Abstract
With the development of the marine economy and the progress of hydroacoustic communication technology in recent years, the demand for marine resources development and protection is increasing, and the demand for in-depth exploration and monitoring of the marine environment is becoming more and more urgent, the underwater acoustic sensor network (UASN) has received widespread attention, and at the same time, the enhancement of marine security and national defense needs has also pushed the development of hydroacoustic sensor technology.
Given the characteristics of long delay, narrow bandwidth, and low rate inherent in hydroacoustic communication, the traditional land-based power allocation methods and signal transmission modes are not applicable to underwater environments, and UASNs mainly rely on acoustic waves as the communication medium. Optimizing the power allocation, choosing the appropriate transmission mode, and finding the equilibrium point of power allocation in the current environment can improve the data transmission rate and make the network more adaptable to the underwater environment. Effective power allocation and transmission modes can improve network performance and reliability to support applications such as underwater monitoring, resource exploration, and communication in the presence of bandwidth constraints, limited energy, and interference problems. Traditional power allocation algorithms are highly affected by the dynamic uncertainty of the UASN environment, and at the same time, each node is not a separate individual. Therefore, optimizing the power allocation and choosing the appropriate power allocation requires the collaboration of multiple nodes.
In this paper, I design the Adaptive UASN Link Power Algorithm (AULPA) to address the UASN performance optimization problem. AULPA effectively reduces the signal interference by adjusting the power of each node, increases the signal co-concurrency among different nodes, and improves the throughput of the network. AULPA targets the dynamic uncertainty of UASN environment and designs a UASN environment model for multi-intelligent body systems. In addition, this paper introduces the implementation details of AULPA under the hydroacoustic sensor network carrier, including the network structure of the intelligences in AULPA, the training process and the parameter tuning method.
Specifically, in the experimental part, a series of experiments are designed to evaluate the total transmission rate, convergence speed, spectral efficiency, and computation time overhead of the network under different transmission modes and parameter settings, and compared with other reinforcement learning optimization methods DQN and DDPG. The experimental results show that the UASN optimization method of AULPA can improve the performance of the network to a certain extent, which makes the network achieve better performance in terms of total transmission rate, convergence speed of the algorithm, and spectral efficiency.
Key words:
Underwater Acoustic Sensor Network; Power; Transmission mode; MADDPG; Channel capacity