Rethinking Power Allocation in OFDMA: A Throughput-Optimized Multi-User Strategy

Modern networks rely on orthogonal frequency division multiple access (OFDMA) to efficiently allocate bandwidth among multiple users. While resource allocation mechanisms primarily focus on bandwidth distribution, transmit power is typically flat along the spectrum, potentially limiting achievable data rates due to heterogeneous channel conditions. In this paper, we investigate the impact of power allocation on modulation and coding scheme (MCS) selection. Specifically, we formulate an optimization problem aimed at maximizing the aggregate data rate in a wireless local area network (WLAN), subject to constraints on minimum packet delivery ratio (PDR). The key insight is that reallocating transmit power across users can alter the signal-to-noise ratio (SNR), enabling higher MCS levels and improving throughput without modifying bandwidth assignments. Due to the discrete nature of MCS levels, the problem is inherently combinatorial and difficult to solve optimally. To address this, we propose a practical heuristic inspired by water-filling principles, combined with a Knapsack-based selection strategy to prioritize users that yield the highest rate gains per unit of power. The approach redistributes power from users with excess margin to those who can benefit from an MCS upgrade. Simulation results based on IEEE 802.11ax settings show that the proposed method improves the aggregate throughput while preserving reliability constraints.