Local search approach for the competitive facility location problem in mobile networks

We present a new leader-follower facility location game for 5G high-speed networks. Two mobile operators, the leader and the follower, compete to attract customers of high-speed internet connection. The leader acts first by opening some base stations, anticipating that the follower will react by creating her own base stations and renting some leader’s stations. Each customer patronizes an operator with the highest speed of connection for him. Each operator aims to maximize own profit. We provide a formulation of this game as a bi-level nonlinear integer programming problem, where the leader and the follower maximize their own profits at the first and second levels, respectively and the customers move from one operator to another until the Nash equilibrium is reached. Local search heuristics for this problem are designed and tested on randomly generated and real data. Computational experiments show high efficiency of the approach.