This study shows a distributed and accurate algorithm for location based on manifold learning algorithms: Hessian Local Linear Embedded (HLLE) algorithm and a localization method framework based on manifold learning is proposed. The method takes advantage of additional information such as estimated distances between neighbors, or reference nodes` positions. The algorithm based on Hessian local linear embedded to obtain the relative map of nodes. Through simulation studies, we demonstrate that the algorithm is more robust to measurement error than previous proposals, especially when nodes are positioned relatively uniformly throughout the plane and it can achieve comparable results using fewer reference nodes than previous methods, if there are no reference nodes, relative coordinates are available.