A parallel algorithm for phase retrieval with dictionary learning

We propose a new formulation for the joint phase retrieval and dictionary learning problem with a reduced number of regularization parameters to be tuned. A parallel algorithm based on the block successive convex approximation framework is developed for the proposed formulation. The performance of the algorithm is evaluated when applied to sparse channel estimation in a multi-antenna random access network. Simulation results on synthetic data show the efficiency of the proposed technique compared to the state-of-the-art method.