A Memory-Efficient Learning Framework for Symbol Level Precoding With Quantized NN Weights
A Memory-Efficient Learning Framework for Symbol Level Precoding With Quantized NN Weights
复制标题
DOI：
10.1109/ojcoms.2023.3285790
复制
发表时间：
2021-10
影响因子：
7.9
通讯作者：
A. Mohammad;C. Masouros;Y. Andreopoulos
中科院分区：
文献类型：
--
作者：
A. Mohammad;C. Masouros;Y. Andreopoulos
提出了一种基于符号级预编码（SLP）的高效记忆深度神经网络框架。我们关注具有现实有限精度权值的深度神经网络，并采用基于无监督深度学习（DL）的SLP模型（SLP- dnet）。我们应用随机量化（SQ）技术得到相应的量子化版本，称为SLP-SQDNet。该方案通过量化DNN权重的可扩展百分比，提供了可扩展的性能与内存权衡，我们探索了二进制和三元量化。我们的研究结果表明，虽然SLP-DNet提供了近乎最佳的性能，但其通过SQ的量化版本分别为基于二进制和基于三元的SLP-SQDNets产生$\sim 3.46\times $和$\sim 2.64\times $模型压缩。我们还发现，与基于SLP优化和SLP- dnet相比，我们的建议分别提供了20倍和10倍的计算复杂度降低。
This paper proposes a memory-efficient deep neural network (DNN) framework-based symbol level precoding (SLP). We focus on a DNN with realistic finite precision weights and adopt an unsupervised deep learning (DL) based SLP model (SLP-DNet). We apply a stochastic quantization (SQ) technique to obtain its corresponding quantized version called SLP-SQDNet. The proposed scheme offers a scalable performance vs memory trade-off, by quantizing a scalable percentage of the DNN weights, and we explore binary and ternary quantizations. Our results show that while SLP-DNet provides near-optimal performance, its quantized versions through SQ yield $\sim 3.46\times $ and $\sim 2.64\times $ model compression for binary-based and ternary-based SLP-SQDNets, respectively. We also find that our proposals offer $\sim 20\times $ and $\sim 10\times $ computational complexity reductions compared to SLP optimization-based and SLP-DNet, respectively.