In recent years, phase retrieval has received much attention in statistics, applied mathematics and optical engineering. In this paper, we propose an efficient algorithm, termed Subspace Phase Retrieval (SPR), which can accurately recover an -dimensional -sparse complex-valued signal given its magnitude-only Gaussian samples if the minimum nonzero entry of satisfies . Furthermore, if the energy sum of the most significant elements in is comparable to , the SPR algorithm can exactly recover with magnitude-only samples, which attains the information-theoretic sampling complexity for sparse phase retrieval. Numerical Experiments demonstrate that the proposed algorithm achieves the state-of-the-art reconstruction performance compared to existing ones.
View on arXiv