让无线信号预测自动识别每条链路的不确定性，提升网络规划精度。
Heteroscedastic Neural Networks for Path Loss Prediction with Link-Specific Uncertainty
- 用神经网络同时预测信号衰减均值和链路特异性方差。
- 测试误差7.4 dB，95%置信区间覆盖率达95.1%，区间平均宽29.6 dB。
- 适合需要精准信号预测的通信系统设计与故障自诊断场景。
传统及现代基于机器学习的路径损耗模型通常假设预测方差恒定。本文提出一种神经网络，通过最小化高斯负对数似然，联合预测均值与链路特异性方差，实现异方差不确定性估计。我们在大型公开射频路测数据集的盲测集上，对比了共享、部分共享与独立参数架构，以准确率、校准度与锐度为评估指标。共享参数架构表现最佳，达到7.4 dB的均方根误差，95%预测区间的覆盖率高达95.1%，平均区间宽度为29.6 dB。这些不确定性估计可支持链路级覆盖率余量设定，提升射频规划与干扰分析效果，并提供模型弱点的有效自诊断能力。
原文摘要 · Abstract (English)
Traditional and modern machine learning-based path loss models typically assume a constant prediction variance. We propose a neural network that jointly predicts the mean and link-specific variance by minimizing a Gaussian negative log-likelihood, enabling heteroscedastic uncertainty estimates. We compare shared, partially shared, and independent-parameter architectures using accuracy, calibration, and sharpness metrics on blind test sets from large public RF drive-test datasets. The shared-parameter architecture performs best, achieving an RMSE of 7.4 dB, 95.1 percent coverage for 95 percent prediction intervals, and a mean interval width of 29.6 dB. These uncertainty estimates further support link-specific coverage margins, improve RF planning and interference analyses, and provide effective self-diagnostics of model weaknesses.
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。