研究目的
To reduce the pilot overhead and improve the channel estimation accuracy in the massive MIMO system, addressing the off-grid problem where true angles are not on discretized grid points.
研究成果
The proposed algorithm effectively addresses the off-grid problem in mm-wave massive MIMO systems with hybrid precoding by using a parallel structure to reduce computational complexity and an iterative grid refinement strategy to improve estimation accuracy. Simulation results show superior performance in terms of NMSE and computational efficiency compared to existing methods.
研究不足
The study is based on simulations and may not account for all real-world channel impairments. The algorithm's performance relies on the initial grid and may have convergence issues in highly noisy environments. Computational complexity, although reduced, could still be high for very large systems.
1：Experimental Design and Method Selection:
The methodology involves a channel estimation algorithm based on sparse signal reconstruction (SSR) and grid refinement. It transforms the 2D joint AoAs/AoDs estimation into two 1D sub-problems to reduce computational complexity. An iterative grid refinement strategy is used to minimize off-grid errors by constructing an objective function from the orthogonality between signal and noise subspaces.
2：Sample Selection and Data Sources:
Simulations are conducted with parameters such as NR = NT = 64 antennas, NRF = 4 RF chains, L = 3 paths, and SNR defined as 10log10E[||x(n)||^2]/σ_n^
3：List of Experimental Equipment and Materials:
No specific physical equipment is mentioned; the study is based on numerical simulations.
4：Experimental Procedures and Operational Workflow:
The algorithm includes steps: calculate initial sparse support set using SSR, refine grid iteratively using closed-form solutions for off-grid errors, estimate AoAs and AoDs separately, and estimate path gains using least squares estimation (LSE).
5：Data Analysis Methods:
Performance is evaluated using normalized mean square error (NMSE) and root mean square error (RMSE) metrics, with comparisons to existing algorithms like ACS, MUSIC, IR, and DL-based methods.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
