1: \begin{abstract}

2: In this study, we propose an \ac{OAC} scheme to calculate the \ac{MV} for \ac{FEEL}. %The proposed scheme aims at distributed learning by the \ac{MV}. 

3: With the proposed approach, \acp{ED} transmit the signs of local stochastic gradients, i.e., votes, by activating one of two orthogonal resources. The \acp{MV} at the \ac{ES} are obtained with non-coherent detectors by exploiting the accumulations on the resources. Hence, the proposed scheme eliminates the need for \ac{CSI} at the \acp{ED} and \ac{ES}. In this study, we analyze various gradient-encoding strategies through the weight functions and waveform configurations over  \ac{OFDM}. We show that specific weight functions that enable absentee \acp{ED} (i.e., \ac{HPA}) or weighted votes (i.e., \ac{SP})  can substantially reduce the probability of detecting the incorrect \ac{MV}. By taking path loss, power control, cell size, and fading channel  into account, we prove the convergence of the distributed learning for a non-convex function for \ac{HPA}.

4: Through simulations, we show that the proposed scheme  with \ac{HPA} and \ac{SP} can provide  high test accuracy even when the time-synchronization and the power control  are not ideal under heterogeneous data distribution scenarios.

5: \end{abstract}

6: