This paper introduced the triangulation method using the relations of the signal magnitudes and distances between each sensor and impact location. In order to extract the useful feature having good correlation with these distances, the time of arrival (TOA) data of sensor signal was applied. Using the signal magnitudes and TOA data, the effective features could be derived for estimating the distances between each sensor and impact location. For investigating this suggested algorithm, low-velocity impact experiments were carried out on the composite stiffened plate as a target structure. The acoustic emission signals were captured by the commercial high-speed optical fiber sensing system. From the impact signals on the baseline locations (30 points), the regression model was constructed for estimating the distances of impact location from each sensor on the target area. Then, through the non-baseline signals (20 points), the validations of constructed regression model were performed. As a result, this regression model is able to accurately predict the AE source locations on the non-baseline cases.