1: \begin{abstract}

2:     Intelligent reflected surfaces (IRSs) are envisioned as a cost-effective solution for achieving high spectrum and energy efficiency in future wireless networks due to its ability to customize the wireless propagation environments. In this paper, we investigate the resource allocation problem for multi-user IRS-assisted multiple-input single-output (MISO) systems. We first formulate the joint optimization of precoding at the base station (BS) and the phase shifts at the IRS to a non-convex optimization problem with objective to minimize the total transmit power. Then, by employing generalized Bender's decomposition approach, we develop an optimal resource allocation algorithm which can produce the global optimal solution to the formulated optimization problem with guaranteed convergence. Simulation results reveal that the proposed algorithm can significantly reduce the BS total transmit power compared to two state-of-the-art suboptimal schemes.

3: \end{abstract}

4: