1: \begin{abstract}

2: Intelligent reflecting surface (IRS) has gained tremendous attention

3: recently as a disruptive technology for beyond 5G networks. In this

4: paper, we consider the problem of secrecy rate maximization for an

5: IRS-assisted Gaussian multiple-input multiple-output multi-antenna-eavesdropper

6: (MIMOME) wiretap channel (WTC). In this context, we aim to jointly

7: optimize the input covariance matrix and the IRS phase shifts to maximize

8: the achievable secrecy rate of the considered system. To solve the

9: formulated problem which is non-convex, we propose an iterative method

10: based on the block successive maximization (BSM), where \emph{each

11: iteration is done in closed form}. More specifically, we maximize

12: \emph{a lower bound} on the achievable secrecy rate to update the

13: input covariance matrix for fixed phase shifts, and then maximize

14: the (exact) achievable secrecy rate to update phase shifts for a given

15: input covariance. We present a convergence proof and the associated

16: complexity analysis of the proposed algorithm. Numerical results are

17: provided to demonstrate the superiority of the proposed method compared

18: to a known solution, and also to show the effect of different parameters

19: of interest on the achievable secrecy rate of the IRS-assisted MIMOME

20: WTC.

21: \end{abstract}

22: