Open AccessHierarchical classification of time series data aggregation in underwater wireless sensor networks
Author(s) -
D. Ruby,
J. Jeyachidra
Publication year - 2020
Publication title -
underwater technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.229
H-Index - 19
eISSN - 1756-0551
pISSN - 1756-0543
DOI - 10.3723/ut.37.053underwater
Subject(s) - underwater , energy consumption , computer science , wireless sensor network , real time computing , time series , node (physics) , similarity (geometry) , energy (signal processing) , data mining , variance (accounting) , submarine , statistics , artificial intelligence , engineering , machine learning , mathematics , computer network , marine engineering , oceanography , accounting , structural engineering , geology , electrical engineering , business , image (mathematics)
Environmental fluctuations are continuous and provide opportunities for further exploration, including the study of overground, as well as underground and submarine, strata. Underwater wireless sensor networks (UWSNs) facilitate the study of ocean-based submarine and marine parameters details and data. Hardware plays a major role in monitoring marine parameters; however, protecting the hardware deployed in water can be difficult. To extend the lifespan of the hardware, the inputs, processing and output cycles may be reduced, thus minimising the consumption of energy and increasing the lifespan of the devices. In the present study, time series similarity check (TSSC) algorithm is applied to the real-time sensed data to identify repeated and duplicated occurrences of data for reduction, and thus improve energy consumption. Hierarchical classification of ANOVA approach (HCAA) applies ANOVA (analysis of variance) statistical analysis model to calculate error analysis for realtime sensed data. To avoid repeated occurrences, the scheduled time to read measurements may be extended, thereby reducing the energy consumption of the node. The shorter time interval of observations leads to a higher error rate with lesser accuracy. TSSC and HCAA data aggregation models help to minimise the error rate and improve accuracy.