A Novel Hybrid Structure of Denoising for ECG-TDM Signals using Wavelet Packet Decomposition

摘要

Multiplexing of signals enables the usage of same channel for multiple signals at the same time. Among multiplexing schemes, Time division multiplexing is one of the fundamental and basic multiplexing scheme. TDM was utilized in popular Tl carrier system, which transfers voice signals in a single channel. TDM divides the total time scale into serial slots, where these slots are used to carry samples of signals in a round robin manner. When the TDM signal is presented to channel, the noise in channel, effects the signal over some time, or some frequency with certain effect on amplitude. Hence, when the TDM signal is attacked by noise, few samples of few or all the base signals may be effected by noise. The solution to this problem was in many forms. But all these forms takes the TDM signal as a base or whole signal as a unit, and run the denoising schemes. But the base signals which constitutes the TDM signal may have variety of properties. These signals will be effected by noise in a different way from each other. Hence towards this end, the main objective and contribution is to perform the denoising after demultiplexing. Here, the denoising of different signals may be done with different denoising schemes or with a variety of quantity the denoising may be applied. The other aspect of this research work is to find out optimal denoising technique for base signals based on base signal properties. The denoising techniques considered are adaptive filtering techniques, wavelet based denoising and wavelet packed based denoising. ECG and Audio signals are considered as base signals in this work. Simulation results suggest that the proposed structure result in good performance.