Non-cooperative power scheduling for wireless MIMO networks

This paper presents a non-cooperative method to maximize the throughput of a set of distributed links in wireless MIMO ad hoc networks. The space-time power scheduling in each link is built into a non-cooperative game where its utility function is identified and maximized. The link quality is introduced into the pricing function in order to obtain Pareto improvement. The gradient projection methods are compared with the proposed game theoretic approach by simulations, and the results show that the proposed method provides a distributed scheduling of resources with low computational complexity and without significantly sacrificing the network throughput.