This paper presents and analyzes a one-shot coding scheme for the \gls{ota} computation over a fast-fading multiple-access wireless channel. The assumed channel model incorporates correlations both in fading and noise over time as well as among users. The model also allows for non-Gaussian components in fading and noise, provided that the distributions are sub-Gaussian (as is the case for a sum of Gaussian and bounded random variables), rendering the proposed scheme robust to a large class of non-Gaussian interference and noise known to occur in many practical scenarios. OTA computation has a huge potential for reducing communication cost in applications such as Machine Learning (ML)-based distributed anomaly detection in large wireless sensor networks. We illustrate this potential through extensive numerical simulations.

中文翻译：

---

相关频道中的无线计算

本文提出并分析了一种用于在快速衰落多址无线信道上计算 \gls{ota} 的一次性编码方案。假设的信道模型结合了衰落和噪声随时间以及用户之间的相关性。该模型还允许衰落和噪声中的非高斯分量，前提是分布是亚高斯分布（就像高斯和有界随机变量的总和的情况），使得所提出的方案对一大类非高斯分量具有鲁棒性。已知在许多实际场景中会发生高斯干扰和噪声。OTA 计算在降低大型无线传感器网络中基于机器学习 (ML) 的分布式异常检测等应用中的通信成本方面具有巨大潜力。我们通过广泛的数值模拟来说明这种潜力。