Worst-Case Design for RIS-Aided Over-the-Air Computation With Imperfect CSI

Over-the-air computation (AirComp) enables fast wireless data aggregation at the receiver through concurrent transmission by sensors in application of Internet-of-Things (IoT). To further improve performance AirComp under unfavorable propagation channel conditions, we consider problem distortion minimization a reconfigurable intelligent surface (RIS)-aided system. In particular, take into account an additive bounded uncertainty state information (CSI) and total power constraint, jointly optimize transceiver (Tx-Rx) RIS phase design from perspective worst-case robustness minimizing mean squared error (MSE) computation. solve this intractable nonconvex problem, develop efficient alternating algorithm where both solutions to robust sub-problem joint Tx-Rx are obtained closed forms. Simulation results demonstrate effectiveness proposed method.