An Adaptive Switching Algorithm for Element Resource Scheduling in Digital Array Radars Based on an Improved Ant Colony Optimization

To address the conflict between real-time performance and optimal resource allocation in large-scale digital array radars, this paper proposes a novel resource scheduling framework that integrates graph-theoretic modeling with an adaptive heuristic strategy. Unlike traditional methods, we formulate the multi-beam scheduling problem as a constrained connected subgraph optimization task. To solve this NP-hard problem, an Improved Ant Colony Optimization (I-ACO) algorithm is designed, incorporating pheromone boundary constraints and elite update strategies to effectively balance exploration and exploitation within complex solution spaces. Furthermore, a load-aware Adaptive Algorithm Switching (AAS) strategy is introduced. This mechanism dynamically transitions between the globally optimized I-ACO and a rapid, utility-guided greedy approach based on real-time system load, effectively resolving the trade-off between solution quality and response speed. Experimental results demonstrate that the proposed method reduces solution costs by up to 23.5% compared to greedy algorithms and increases the scheduling success rate to 99.2% under high-load conditions, while significantly improving long-term system load balancing by 41.5%.