1: \begin{abstract}

2: % Analog over-the-air federated edge learning is anticipated to play a crucial role in realizing large-scale, low-latency intelligent wireless networks.

3: Leveraging over-the-air computations for model aggregation is an effective approach to cope with the communication bottleneck in federated edge learning. 

4: By exploiting the superposition properties of multi-access channels, this approach facilitates an integrated design of communication and computation, thereby enhancing system privacy while reducing implementation costs.

5: However, the inherent electromagnetic interference in radio channels often exhibits heavy-tailed distributions, giving rise to exceptionally strong noise in globally aggregated gradients that can significantly deteriorate the training performance. 

6: To address this issue, we propose a novel gradient clipping method, termed Median Anchored Clipping (MAC), to combat the detrimental effects of heavy-tailed noise. 

7: We also derive analytical expressions for the convergence rate of model training with analog over-the-air federated learning under MAC, which quantitatively demonstrates the effect of MAC on training performance. 

8: Extensive experimental results show that the proposed MAC algorithm effectively mitigates the impact of heavy-tailed noise, hence substantially enhancing system robustness. 

9: % our proposed MAC algorithm significantly enhances system robustness, effectively mitigating the impacts of heavy-tailed noise in analog over-the-air federated learning.

10: \end{abstract}

11: