This research proposes the shape reconstruction algorithm for the deformation monitoring of composite structures using the high-resolution distributed strain data, which is obtained by the embedded optical fiber network. The accurate shape monitoring system is expected to be useful for full-scaled structural monitoring of larger composite structures. Once the optical fiber network is installed in the structure, such global shape monitoring system will provide an excellent tool for many monitoring applications in their life time. In this paper, we constructed the reconstruction algorithm for the deflection in the bending deformation of composite laminates. The high-resolution distributed strain data was obtained by one of the optical fiber sensing systems, pulse-pre-pump Brillouin optical time domain analysis (PPP-BOTDA) system. We fabricated a composite laminate specimen with an embedded optical fiber network, and the cantilever bending test was carried out. Using obtained distributed data, the deflection of the specimen was predicted using the constructed algorithm. The deflections were successfully reconstructed with a few percents of the prediction error. From the result, it was shown that the deformation of composite structures was able to be reconstructed with high estimation accuracy using our algorithm, which uses distributed strain data obtained by the embedded optical fiber network.