This repo contains the development of the final assignment from the Operations Research course at UFMG (ministred by Prof. Eduardo Carrano).
This work proposes a solution to the WiFi router allocation problem, optimizing the installation cost. The problem is modeled to minimize the cost with routers, respecting the constraint of a minimum signal power value at all receiving points. The formulation was tested for some created instances and the result was satisfactory, both in qualitative terms and in computational feasibility for instances with a higher number of variables.
Following this link, you'll see the python notebook used to implement the mathemathical model of optimization.
Following this link, you can access the paper generated of the work.
Thanks, Vittor