1: \begin{abstract}

2: % v2

3: We propose a novel uplink communication method, coined \emph{random orthogonalization}, for federated learning (FL) in a massive multiple-input and multiple-output (MIMO) wireless system. The key novelty of random orthogonalization comes from the tight coupling of FL model aggregation and two unique characteristics of massive MIMO -- channel hardening and favorable propagation. As a result, random orthogonalization can achieve natural over-the-air model aggregation without requiring transmitter side channel state information, while significantly reducing the channel estimation overhead at the receiver. Theoretical analyses with respect to both communication and machine learning performances are carried out. In particular, an explicit relationship among the convergence rate, the number of clients and the number of antennas is established. Experimental results validate the effectiveness and efficiency of random orthogonalization for FL in massive MIMO.

4: 

5: 

6: % v1

7: %In this work, we put forward a novel framework for federated learning (FL) in massive multiple-input and multiple-output (MIMO) wireless systems. By investigating random orthogonalization properties of massive MIMO wireless channels, we propose a simple-yet-effective wireless FL protocol that requires little channel estimation overhead and achieves low system latency. To evaluate the performance of the proposed scheme, we derive the Cramer-Rao lower bounds (CRLBs) of estimates of FL parameters. Furthermore, taking the communication interference and noise into account, a novel convergence bound is derived for the proposed scheme over massive MIMO systems. Especially, we establish an explicit relationship of the convergence rate between the number of clients and the number of antennas. The numerical results validate the effectiveness of the proposed scheme.

8: \end{abstract}

9: