A Pareto meta-front approach to the optimization of reconfigurable millimeter-wave passive reflectors

Abstract

The paper proposes a Pareto meta-front approach for the multi-objective design of reconfigurable millimeter-wave passive reflectors. For each value of a generalized cost parameter, such as wave frequency, a conventional Pareto front is generated by simultane-ously maximizing the Radar Cross Section (RCS) and minimizing the physical surface area of the array. The utopia point is then determined, and the Pareto-optimal solution with the minimum Euclidean distance from it is selected as the best compromise. Repeating this procedure over the parameter range yields the Pareto meta-front, defined as the locus in the objective space that tracks the evolution of the preferred trade-off across a family of Pareto fronts indexed by different values of the generalized cost parameter. The solutions associ-ated with points on this curve are referred to as meta-optimal solutions. The proposed method is applied to a reconfigurable passive dipole array for intelligent reflecting surface applications. The array employs variable element lengths and switchable reactive loads to control the phase distribution and electronically steer the reflected beam. Array geometry and load values are jointly optimized over the 50–70 GHz frequency range to maximize the RCS while minimizing reflector area and maintaining performance over a broad angular range. In the proposed frequency-indexed formulation, the incident-wave frequency serves as the generalized cost parameter. The resulting meta-front identifies frequency-dependent configurations that provide an effective compromise between reflectivity and physical size, establishing a systematic frequency-indexed multi-objective framework for wideband elec tromagnetic wave redirection.