Download
The solver performs an exhaustive search on the space defined using simple formal language grammar constructs to identify solutions that meet the objective and constraints. It can be used to easily define complex discrete multi-dimensional combinatorial problems that are otherwise difficult to model using Mixed Integer Linear Programming solvers. The solver does not make any assumptions on continuity of the search space. The solver is implemented in PHP. A GoJS based visualization is provided for model's grammar.
Two system design model samples are provided. One uses the solver to identify low power network design for Passive Optical Networks given a set of choices for provider and customer premises equipment and a set of optical domain constraints. Second uses the solver to identify a low power switch design for data center networks given a set of choices for various sub-components within a switch and a set of constraints.
Solver uses PHP 5.6 and GoJS from http://gojs.net
Features
- Constraint based discrete optimization
- Formal language grammar for model definition
- Exhaustive search with branch and bound
- Visualization using GoJS
Project Samples
