Date of Award


Document Type


Degree Name

Master of Science


Department of Operational Sciences

First Advisor

Matthew J. Robbins, PhD.


The United States Army currently employs a shoot-shoot-look firing policy for air defense. As the Army moves to a networked defense-in-depth strategy, this policy will not provide optimal results for managing interceptor inventories in a conflict to minimize the damage to defended assets. The objective for air and missile defense is to identify the firing policy for interceptor allocation that minimizes expected total cost of damage to defended assets. This dynamic weapon target assignment problem is formulated first as a Markov decision process (MDP) and then approximate dynamic programming (ADP) is used to solve problem instances based on a representative scenario. Least squares policy evaluation (LSPE) and least squares temporal difference (LSTD) algorithms are employed to determine the best approximate policies possible. An experimental design is conducted to investigate problem features such as conflict duration, attacker and defender weapon sophistication, and defended asset values. The LSPE and LSTD algorithm results are compared to two benchmark policies (e.g., firing one or two interceptors at each incoming tactical ballistic missile (TBM)). Results indicate that ADP policies outperform baseline polices when conflict duration is short and attacker weapons are sophisticated. Results also indicate that firing one interceptor at each TBM (regardless of inventory status) outperforms the tested ADP policies when conflict duration is long and attacker weapons are less sophisticated.

AFIT Designator


DTIC Accession Number