Energy-Efficient Over-the-Air Computation for Relay-Assisted IoT Networks

Over-the-air computation (AirComp) utilizes the superposition of wireless multiple access channels (MAC) to achieve efficient data aggregation, which is a very effective way to achieve large-scale data transmission and edge intelligence in the intelligent Internet of Things (IoT). In this letter, we propose an amplify and forward (AF) relay-assisted two-hop AirComp system, where the AF relay is used for forwarding signals from sensors. We present the problem of minimizing the sum power (SP) of the sensors under the constraints of available relay transmit power and data aggregating mean square error (MSE). The original problem includes optimizing transmitter (Tx) scaling factor for the sensors, optimizing Tx scaling factor for the relay, and optimizing receiver (Rx) scaling factor for the fusion center (FC). To address the non-convexity of the objective function and constraints of the optimization problem, we transform the original problem into two sub-problems and solve it by iterative optimization algorithm, and each subproblem can be solved with a closed-form optimal solution. In simulations, our proposed SP minimization scheme outperforms benchmark schemes with different parameter settings.