Augmenting SAT Solvers for Network Configuration/Planning
Sharad Malik · 2006
This project explored the possibility of alternate encodings of the planning problem and extensions to satisfiability (SAT) solvers that can better capture the constraints and objectives for network configurations. For example, the ability to directly deal with arithmetic constraints dealing with configuration cost, or probabilistic failure modes may lead to more compact encodings that are then dealt with more sophisticated decision procedures. The end goal is to capture the constraints and objectives of network planning in a decision problem and then solve these using efficient decision procedures in a scalable way.