1: \begin{abstract}

2: Phase retrieval has been an attractive but difficult problem rising from physical science,

3: and there has been a gap between state-of-the-art theoretical convergence analyses and the corresponding efficient retrieval methods.

4: Firstly, these analyses all assume that the sensing vectors and the iterative updates are independent, 

5: which only fits the ideal model with infinite measurements but not the reality, where data are limited and have to be reused.

6: Secondly, the empirical results of some efficient methods, such as the randomized Kaczmarz method, 

7: show linear convergence, which is beyond existing theoretical explanations considering its randomness and reuse of data.

8: In this work, we study for the first time, without the independence assumption, the convergence behavior of the randomized Kaczmarz method for phase retrieval.

9: Specifically, beginning from taking expectation of the squared estimation error with respect to the index of measurement by fixing the sensing vector and the error in the previous step,

10: we discard the independence assumption, rigorously derive the upper and lower bounds of the reduction of the mean squared error, 

11: and prove the linear convergence.

12: This work fills the gap between a fast converging algorithm and its theoretical understanding.

13: The proposed methodology may contribute to the study of other iterative algorithms for phase retrieval and other problems in the broad area of signal processing and machine learning.

14: 

15: {\bf Keywords:} Phase retrieval algorithm, performance analysis, independence assumption, linear convergence rate, data reuse

16: \end{abstract}

17: