Mining asynchronous coincidence pattern is a difficult task in multi-data streams. The main contributions of this work included: (1) The filter technique of Haar Wavelet is investigated and applied to mining asynchronous coincidence pattern in multi-streams; (2) The Wavelet coefficient series are applied to the measurement of asynchronous coincidence between data streams. A series of theorems are proved to ensure the validity of measuring asynchronous coincidence; (3) The anti-noise increment algorithms are designed on loop sliding windows to mine asynchronous coincidence pattern and implemented with complexity O(n2); (4) The extensive experiments on real data are given to validate algorithms.