LAUSR

In future engineered systems for medical applications, a tight real-time integration between physical and computational processes will be required. That integration is achieved using feedback control loops which need high quality input data streams. However, hardware platforms can barely provide such high-quality data sequences (especially if mobile nodes are considered), and mechanisms to improve and polish physical and biological signals are then necessary. This paper proposes a predictor-corrector algorithm to improve the quality and precision of data (biological) signals in Internet of Medical Things deployments, especially if composed of mobile nodes. The proposed algorithm employs an Artificial Intelligence approach and statistical learning techniques to predict future data samples and correct errors in received information. Employed mathematical models follow a prediction-correction scheme and are based on complex functions, Laurent series and the idea of complex envelope. Simulation techniques are used to evaluate the performance of the proposed solution, showing that it improves the precision of traditional linear interpolation techniques up to 85%, and cubic splines up to 20%. Processing delay during operation is, for the referred precision, around 200ms.