This article presents limited feedback-based precoder quantization schemes for Interference Alignment (IA) with bounded channel state information (CSI) uncertainty. Initially, this work generalizes the min-max mean squared error (MSE) framework, followed by the development of robust precoder and decoder designs based on worst case MSE minimization. The proposed precoder and decoder designs capture the effect of CSI uncertainty using a single parameter, which is independent of the CSI uncertainty in the direct links. The IA algorithms derived employing these proposed designs are shown to be globally convergent under certain conditions. Moreover, precoder quantization schemes are presented for scenarios with and without CSI uncertainty for practical implementation of these techniques in systems with limited feedback. An optimal bit allocation scheme is presented to maximize the sum rate via analysis of the rate loss upper bound. Simulation results demonstrate the improved performance of the proposed IA schemes for various scenarios considering imperfect CSI as well as limited feedback.

中文翻译：

---

用于具有有限 CSI 不确定性的鲁棒干扰对准的预编码器反馈方案


本文提出了基于有限反馈的预编码器量化方案，用于具有有限信道状态信息 (CSI) 不确定性的干扰对齐 (IA)。最初，这项工作概括了最小-最大均方误差 (MSE) 框架，随后开发了基于最坏情况 MSE 最小化的鲁棒预编码器和解码器设计。所提出的预编码器和解码器设计使用单个参数捕获CSI不确定性的影响，该参数独立于直接链路中的CSI不确定性。采用这些提出的设计导出的 IA 算法在某些条件下被证明是全局收敛的。此外，针对具有和不具有 CSI 不确定性的场景提出了预编码器量化方案，以便在反馈有限的系统中实际实现这些技术。通过对速率损失上限的分析，提出了一种优化比特分配方案以最大化总速率。仿真结果表明，考虑到不完善的 CSI 以及有限的反馈，所提出的 IA 方案在各种场景下的性能都有所提高。