提出新型自适应量化方法，提升边缘端MIMO检测模型精度与效率
Deep Unfolding with Kernel-based Quantization in MIMO Detection
- 基于核密度估计与最大均值差异，动态对齐高低精度激活分布
- 在不同信道条件下自适应调整量化步长，避免性能下降
- 适用于资源受限的边缘计算场景，显著降低推理延迟
边缘计算的发展对多输入多输出（MIMO）检测任务的高效能模型部署提出了严峻挑战。将深度展开网络如PGD-Nets和ADMM-Nets部署到资源受限的边缘设备时，采用量化方法存在困难。现有基于量化感知训练（QAT）的量化方法因依赖激活分布的参数假设和固定量化步长，导致性能下降。为此，本文提出一种新的基于核的自适应量化（KAQ）框架。通过联合使用核密度估计（KDE）与最大均值差异（MMD），实现全精度与量化模型间激活分布的对齐，无需预先分布假设。此外，引入动态步长更新机制，根据无线网络的信道条件自适应调整量化步长。大量仿真表明，所提KAQ框架在准确率上优于传统方法，并成功降低了模型推理延迟。
原文摘要 · Abstract (English)
The development of edge computing places critical demands on energy-efficient model deployment for multiple-input multiple-output (MIMO) detection tasks. Deploying deep unfolding models such as PGD-Nets and ADMM-Nets into resource-constrained edge devices using quantization methods is challenging. Existing quantization methods based on quantization aware training (QAT) suffer from performance degradation due to their reliance on parametric distribution assumption of activations and static quantization step sizes. To address these challenges, this paper proposes a novel kernel-based adaptive quantization (KAQ) framework for deep unfolding networks. By utilizing a joint kernel density estimation (KDE) and maximum mean discrepancy (MMD) approach to align activation distributions between full-precision and quantized models, the need for prior distribution assumptions is eliminated. Additionally, a dynamic step size updating method is introduced to adjust the quantization step size based on the channel conditions of wireless networks. Extensive simulations demonstrate that the accuracy of proposed KAQ framework outperforms traditional methods and successfully reduces the model's inference latency.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。