Integrated Sensing, Communication, and Computation Over-the-Air: Beampattern Design

In this paper, an integrated sensing, communication, and computation over-the-air (ISCCO) architecture is developed, where sensors perform radar sensing, information communication and computation at the same spectrum by combining over-the-air computation (AirComp) with dual-functional radar-communication techniques. Under the omni-directional and directional radar, beampattern design problems are formulated aiming at minimizing the AirComp error. A tradeoff beampattern design is further investigated to balance the performance of radar sensing and AirComp. To deal with these non-convex problems, the alternating optimization approach has been proposed, where the optimal beampatterns of radars and the access point are obtained by the projected gradient descent method. Simulations are further conducted to illustrate the performances of the proposed beampattern designs.