Adaptive delta modulation method for arbitrary adaptation coefficients in the sensor network of the mobile ECG monitoring

Abstract

The usage of difference methods for information supply in wireless sensor networks can lead to reduction in their power consumption. However, the usage of adaptive delta modulation (ADM) leads to distortion of processed signals. Therefore, the development of ADM methods and researching their parameters to increase the accuracy of the reconstructed signal from the differential code is an important scientific task. Menshikov G.G. theoretically substantiated the ADM method, which allows minimizing the absolute error in the nth sampling interval. The adaptation coefficient of this method is equal to 2, which is relevant with its convenience for hardware implementation. However, the optimality of its value for reducing the quantization error of the adaptive delta encoder was not considered. Therefore, there was a need to develop an ADM method based on the Menshikov’s ADM for arbitrary adaptation coefficients. The right selection of adaptation coefficients can lead to an increase of signal accuracy which is reproduced based on the received delta code. The article presents a method for arbitrary adaptation coefficients which provides a minimum absolute error on the n^th sampling interval. This approach leads to the smallest RMS error of differential signal from the one in PCM format, and in some cases, it allows reducing the resolution of differential signal without loss of its accuracy. The article also provides the results of computer modeling. A comparative analysis of differential signals during ECG supply helped by ADM with instant adaptation, Winkler’s HIDM, the Menshikov’s method and the developed method for arbitrary adaptation coefficients has been carried out. The proposed ADM method makes it possible to select the adaptation coefficients taking into account the forms of the processed signal.