A graph-based approach for distributed parameter coordination in wireless communication networks

This paper addresses distributed parameters coordination methods for wireless communication systems founded on message-passing algorithms on graphs. This work provides evaluations on the potential of such an approach in a wireless communication network, and compares its performance and convergence properties with those of a baseline selfish/greedy approach. Simulation results for an example application, i.e. frequency reuse planning, is presented and discussed. The results show that graph-based techniques generally have higher probability of reaching (near) optimal solution than the greedy approach, though its rate of convergence tends to be slower and the message size is typically larger. Methods of improving the rate of convergence of graph-based distributed coordination techniques and reducing the associated message size are therefore important topics for future studies.