Using a Multiobjective Approach to Balance Mission and Network Goals Within a Delay Tolerant Network Topology

This thesis investigates how to incorporate aspects of an Air Tasking Order (ATO), a Communications Tasking Order (CTO), and a Network Tasking Order (NTO) within a cognitive network framework. This was done in an effort to aid the commander and or network operator by providing automation for battlespace management to improve response time and potential inconsistent problem resolution. In particular, autonomous weapon systems such as unmanned aerial vehicles (UAVs) were the focus of this research This work implemented a simple cognitive process by incorporating aspects of behavior based robotic control principles to solve the multi-objective optimization problem of balancing both network and mission goals. The cognitive process consisted of both a multi-move look ahead component, in which the future outcomes of decisions were estimated, and a subsumption decision making architecture in which these decision-outcome pairs were selected so they co-optimized the dual goals. This was tested within a novel Air force mission scenario consisting of a UAV surveillance mission within a delay tolerant network (DTN) topology. This scenario used a team of small scale UAVs (operating as a team but each running the cognitive process independently) to balance the mission goal of maintaining maximum overall UAV time-on-target and the network goal of minimizing the packet end-to-end delays experienced within the DTN. The testing was accomplished within a MATLAB discrete event simulation. The results indicated that this proposed approach could successfully simultaneously improve both goals as the network goal improved 52% and the mission goal improved by approximately 6%.