Static Weapon Target Assignment Based on Battle Probabilities and Time Discounted Reward

Abstract

Target-based weapon-target assignment (WTA) aims to minimize the total value of enemies. It means that maximizing the total reduced value of the enemies is the objective of the target-based WTA. The reward of an assignment is typically set as the reduction in the enemy’s value when an ally and an enemy have combat, and the value is calculated by multiplying the current value of the enemy by the probability of the enemy’s survival after the combat. However, allies may be assigned to enemies who are far away if the reward is calculated similarly. Additionally, a method of calculating battle probability that reflects the characteristics and deployment of enemies and allies is needed in order to apply it in the defense industry. In this paper, we propose a target-based static weapon-target assignment to solve these problems. First, we propose a method to calculate battle probabilities for one-to-one, one-to-many, many-to-one, and many-to-many combat. The probabilities are composed of 4 cases; ally-survival-enemy-survival, ally-survival-enemy-destroyed, ally-destroyed-enemy-survival, and ally-destroyed-enemy-destroyed. Then a time-discounted reward for assignment based on the battle probabilities is calculated to consider the time it takes to have combat. Finally, the tank combat simulation results are discussed. The performance of the proposed WTA algorithm is highlighted through an analysis of assignment results and a comparison of outcomes based on the application of time-discounted rewards.