Distributed Attacker Assignment and Interception Control in Multi-Agent Target-Attacker-Defender Games
Year: 2026
Type: preprint
Abstract: This paper addresses the Target-Attacker-Defender (TAD) game within a planar environment, focusing on the protection of a stationary target against incoming threats. To compensate for potential speed disadvantages of the defenders, we leverage geometric pre-positioning within an annular defense perimeter. The problem is formulated as an area interception problem, for which we propose a distributed multi-agent algorithm consisting of two primary modules: target allocation and interception control. For target allocation, an equitable region partitioning strategy is employed to divide the defense area into annular sectors and assign defenders to attackers. Under specified angular-separation and parameter conditions, the proposed partitioning mechanism ensures that each defender is assigned at most one attacker, thereby reducing the complexity of multi-agent assignment. For the control layer, movement protocols based on zero-sum differential games and coverage optimization are designed, achieving Nash equilibrium strategies while maintaining effective barrier coverage. Numerical simulations show that the proposed method enables slower defenders to intercept faster attackers through proactive deployment and maintains nearly constant local computation time as the swarm size increases.
Source: arXiv (Cornell University)
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7223661617