Optimal Distributed Power Allocation for Decode-and-Forward Relay Networks

This paper presents a fully distributed power allocation al gorithm for decode-and-forward (DF) relay networks with a large number of sources, relays, and destination nodes. The well known mathematical decomposition based distributed optimization techniques cannot directly be applied to DF relay networks, because the achievable rate of DF relaying is not strictly co ncave, and thus the local power allocation subproblem may have non-unique solutions. We resolve this non-strict concavity problem by using the idea of proximal point method, which adds some quadratic terms to make the objective function strictly concave. While traditional proximal point methods require a two-layer nested iteration structure, our proposed algorithm has a single-layer iteration structure , which is desirable for on-line implementation. Moreover, our algorithm only needs local information exchange among the source, relay, and destination nodes of each DF relay link, and can easily adapt to variations of network size and topology. In this paper, we establish the convergence and optimality of our fully distributed single-layer iterative algorithm. Numerical results are provided to illustrate the benefits of our proposed algorithm. Index terms− Decode-and-forward, distributed power allocation, wireless relay network.