论文标题
有监督的对比度CSI表示学习，用于大规模的MIMO定位
Supervised Contrastive CSI Representation Learning for Massive MIMO Positioning
论文作者
论文摘要
相似性度量对于使用通道状态信息〜（CSI）的大规模MIMO定位至关重要。在这封信中，我们通过深层卷积神经网络（DCNN）和对比度学习提出了一种新颖的大型MIMO CSI相似性学习方法。考虑了从训练数据集中绘制的多个正面和负CSI样本，设计了对比损失函数。使用损失对DCNN编码进行了训练，以便将正样品映射到锚定编码附近的点，而负样本的编码则远离表示空间中的锚固剂。基于指纹的定位在现实世界中的CSI数据集上的评估结果表明，与其他已知的最新方法相比，学习的相似性指标可显着提高定位精度。
Similarity metric is crucial for massive MIMO positioning utilizing channel state information~(CSI). In this letter, we propose a novel massive MIMO CSI similarity learning method via deep convolutional neural network~(DCNN) and contrastive learning. A contrastive loss function is designed considering multiple positive and negative CSI samples drawn from a training dataset. The DCNN encoder is trained using the loss so that positive samples are mapped to points close to the anchor's encoding, while encodings of negative samples are kept away from the anchor's in the representation space. Evaluation results of fingerprint-based positioning on a real-world CSI dataset show that the learned similarity metric improves positioning accuracy significantly compared with other known state-of-the-art methods.
