During a percutaneous soft tissue puncture, a puncture needle is inserted into the layered tissues with various material properties, and the interactive force at the needle tip occurs discriminately. To measure the force and identify the boundaries of the layered tissues in the puncture path, a fiber optic force sensor was fabricated, and the force data acquired with the sensor were analyzed. In terms of a Fabry-Perot interferometer, the measuring principle of the fiber optic sensor was presented, and its optical link, its integration into the needle and calibration approach were introduced. Then, based on the insertion force signal analysis, the force signals were decomposed via the Mallat algorithm into wavelets, and their patterns were recognized and used to identify the layered tissue boundaries. Finally, needle insertion tests were performed with the swine liver and belly tissue phantoms. The experimental results show that the working range of the force sensor is 0-3.20 N and the measuring resolution is under 0.01 N. Moreover, the insertion test results indicate that the Mallat algorithm is effective to the discrimination of layered tissues. It concludes that the fabricated fiber optic sensor meets the requirements of working range, accuracy and reliability to measure an insertion force. The proposed wavelet transform algorithm is able to identify the boundaries of the layered tissues and is expected to used in the percutaneous control by a robot.