Simple Iterative Method for Blind Carrier Frequency Estimation in Multi-user OFDMA Systems

In recent literature, various blind carrier frequency offset (CFO) estimation algorithms have been developed for Orthogonal Frequency-Division Multiple Access (OFDMA) systems. Nonetheless, the majority of the developed algorithms require signal/noise subspaces for CFO estimation tasks through singular value decomposition (SVD) operations, resulting in high computational complexities. In frequency-selective fading channels, channel conditions and CFOs vary with symbol durations. The existing developed algorithms are inefficient for estimating CFOs on a per-symbol basis. Furthermore, these algorithms often struggle to maintain high MSE performance while reducing complexity. Therefore, this paper introduces a novel adaptive CFO estimation scheme that significantly outperforms traditional methods. A subspace tracking method, featuring a new, fast, and stable implementation of the data projection method (DPM), is employed to dramatically reduce computational complexity. Additionally, a novel tracking scheme is derived between symbols to estimate drifting CFOs during frame duration. Simulation results show that while maintaining a lower computational burden, the proposed scheme achieves MSE performance levels that are competitive with high-resolution grid search methods and superior to ESPRIT-based and other MUSIC-based algorithms. These designs effectively reduce computational complexity, conserving mobile energy, and are particularly suitable for time-varying channels with CFO drifting environments. This demonstrates the clear advantages and practical applicability of the proposed solution in dynamic conditions.