Argumentation-based Negotiation for Automated Sensor Tasking (ANAST)

Traditional sensor management problems rely on planning and scheduling algorithms that formulate solutions based on a predefined set of goals and constraints. Optimization of sensor assignments with static goals and constraints is a well-developed area of research in objective function optimization and planning. However, when the set of goals and constraints are dynamic, as is the case with military surveillance and targeting missions, traditional techniques fail to appropriately address the dynamic nature of the environment in which they operate. A proper dynamic sensor management system requires an adaptive approach to optimization that dynamically allocates sensors and their supporting resources in response to changing goals and constraints. Here, we propose a system for Argumentation-based Negotiation for Automated Sensor Tasking (ANAST) to meet the challenge of this real-time resource allocation problem. This approach applies solutions from argumentation theory by creating a dialogue between multiple user-specified needs, which compete for sensor information and computational resources. Intelligent agents argue on behalf of each task and interact based on a negotiation mechanism that determines the means by which agents may resolve conflicting goals. Negotiation works continuously, providing dynamic adaptation to changes in the mission environment.