The rapid growth in drone air traffic calls for enhanced radar surveillance systems to ensure reliable detection in challenging conditions. Increasing radar scattering cross-section can greatly improve detection reliability in civilian applications. Here, we introduce a concept of evolutionarily designed metamaterials in the form of multilayer stacks of arrays, featuring strongly coupled electric and magnetic resonators. These structures demonstrate a broadband end-fire scattering cross-section exceeding 1 m² at 10 GHz and, despite their compact footprint, achieve over 10% fractional bandwidth, meeting essential radar requirements for high-range resolution. While scattering cross-section and bandwidth are typically contradictory in resonant structures, this trend is circumvented by applying the resonance cascading principle, wherein a series of closely spaced, spectrally aligned resonant multipoles create a coherent response. The resonance cascading is engineered with the aid of multi-objective optimization, implemented on top of a genetic algorithm, operating in a large search space, encompassing over 100 independent variables. Experimentally realized parameters match typical scattering cross-sections of large airborne targets. Consequently, these performance characteristics enable the exploration of highly scattering structures as identifiers for small airborne targets, supporting effective radar-based air traffic monitoring in civilian applications, which we demonstrate through outdoor experiments using the DJI Mini 2 drone.