Improving alignment of noisy signals using an iterative Zero Phase method

This paper deals with the study of the performance of different methods to align noisy one-dimensional signals shifted in time. The methods studied in this paper are the Maximum Position method, the Cross-correlation method and the Zero Phase method. Results show the best performance of the Zero Phase method for middle and high values of SNR. A statistical study is carried out in order to achieve a theoretical expression for the value of the performance of the Zero Phase method. This expression demonstrate the dependence of the accuracy of the method on the number of samples. So, a new method is proposed. First the Zero Phase algorithm is used to align the signals. Then, those samples of the aligned signals which only contain noise are eliminated. At last, the Zero Phase algorithm is again applied over the length-reduced profiles. Results show an improvement up to 48%.