Hierarchical Alignment of Multiple Time Series With Missing Timestamps

Collecting data using a UAV nomadic gateway, which serves as a "go-between" edge device between resource-limited end sensor devices and a cloud instance may lead to data quality issues. Power outages on end devices together with MCU lacking the system clock may not only make collected time series data incomplete but also disrupt timestamping of sequences of samples. The proposed data fusion method performed at the cloud side combines information from a number of sensors to align the sequences along time.