Radio network design using coarse-grained parallel genetic algorithms with different neighbor topology

Choosing transmitter locations among alternatives optimally in a radio network, regarding a known area, to guarantee a stipulated quality of service (QoS), is tackled by coarse-grained parallel genetic algorithms, which maximize the coverage while reducing the number of utilized transmitters. An exclusive local search operator is raised, and the impacts of neighbor topologies are compared. Simulations on a dedicated cluster demonstrate that our local search operator is very effective, in the meantime, increasing the neighboring number of subpopulations will ameliorate optimization quality.