Algorithm 3 Dynamic Time-Warping Position-Fingerprint-Matching Algorithm

Input: fingerprint time series, triangular region, start point, end point, fingerprint map table;  Output: the most similar candidate trajectory within the triangular region;   1. {   2. Fit fingerprint time series by MLS;  3. For (each candidate trajectory satisfying the three conditions mentioned above)  4. Calculate $D_{DTW}$ between the candidate trajectory and the fitted fingerprint time series;  5. Find the minimal $D_{DTW}$, which denotes the most similar candidate trajectory.  6. }