A two-stage bidirectional predictive data acquisition method based on underwater edge devices

Abstract

A two-level bi-directional predictive data acquisition method based on underwater edge devices. By transferring computing from the centralized cloud layer to the distributed edge layer, a bottom-edge-cloud three-layer data collection model is constructed. Utilize the mobility and computing power of underwater edge devices to analyze and predict data based on historical information, so as to effectively reduce acoustic wave communication and sensor energy consumption while ensuring accurate data transmission. Considering the differentiation capabilities of heterogeneous devices, different prediction algorithms are used for the edge layer and bottom layer nodes, and the underwater data collection process is transformed into a two-level bidirectional data prediction process. The invention can be applied in the data collection mechanism of the underwater sensor network.

Description

technical field [0001] The invention relates to the field of data collection in underwater sensor networks, in particular to a two-stage bidirectional predictive data acquisition method based on underwater edge devices. Background technique [0002] In recent years, due to the wide application of underwater wireless sensor networks in marine energy survey, environmental index detection, military surveillance and marine disaster event monitoring, it has received extensive attention at home and abroad. The technologies of underwater surveying and other applications are advancing with the times, and the types of data detected are becoming more and more complicated, such as underwater temperature, salinity, oxygen content, pH, CO 2 , P and other elemental substance concentrations, the collected data values ​​are lengthy and complex. The cost of battery replacement or charging for sensor nodes at the bottom of the ocean is high, and the underwater acoustic wave transmission rate...

AI Technical Summary

Problems solved by technology

First, the general prediction model is generally applied to land wireless sensor networks, and there are few researchers who combine prediction algorithms with underwater sensor networks. Therefore, it is impossible to analyze the suitable model, select the appropriate forecasting algorithm

Second, there is still no good solution to the problems of weak computing power, small storage energy and inability to deliver data to the cloud in time for underwater nodes

Images