On the maximum achievable rates in wireless meshed networks: centralized versus decentralized solutions

In this work we provide the optimal coding strategy for meshed wireless networks, where more links are active simultaneously, assuming as optimality criterion the rates of all the links. We formulate the rate maximization problem as a multi-objective optimization problem (MOP). Assuming a multi-carrier modulation for each user, we show how to allocate the power of each user optimally according to a centralized power distribution algorithm. We also propose a decentralized (suboptimal) but simpler algorithm, based on the idea of Nash equilibrium (NE). Finally, we compare the two strategies showing that the loss, in terms of information rate, of the decentralized strategy based on the iterative water-filling algorithm can be very small with respect to the optimal centralized solution, as the distance between the interfering links is just a few times the distance of each link, thus making the decentralized approach a viable solution.