Performance Optimization for Movable Antenna Enhanced MISO-OFDM Systems

Movable antenna (MA) technology offers a flexible approach to enhancing wireless channel conditions by adjusting antenna positions within a designated region. While most existing works focus on narrowband MA systems, this paper investigates MA position optimization for an MA-enhanced multiple-input single-output (MISO) orthogonal frequency-division multiplexing (OFDM) system. This problem appears to be particularly challenging due to the frequency-flat nature of MA positioning, which should accommodate the channel conditions across different subcarriers. To overcome this challenge, we discretize the movement region into a multitude of sampling points, thereby converting the continuous position optimization problem into a discrete point selection problem. Although this problem is combinatorial, we develop an efficient partial enumeration algorithm to find the optimal solution using a branch-and-bound framework, where a graph-theoretic method is incorporated to effectively prune suboptimal solutions. In the low signal-to-noise ratio (SNR) regime, a simplified graph-based algorithm is also proposed to obtain the optimal MA positions without the need for enumeration. Simulation results reveal that the proposed algorithm outperforms conventional fixed-position antennas (FPAs), while narrowband-based antenna position optimization can achieve near-optimal performance.