Utility-based cross layer optimization for OFDMA systems using the β-min-sum belief propagation algorithm

This paper investigates the problem of resource allocation and scheduling for an orthogonal frequency division multiple access (OFDMA) broadband wireless network. We formulate a utility-based cross-layer resource management framework that exploits instantaneous channel state information at the base station in order to schedule traffic during each frame. During each decision epoch, the objective is to find an efficient and fair rate and sub-carrier allocation policy that maximizes the average network utility subject to certain constraints. We develop an equivalent graph theoretical model for the cross-layer assignment problem and show its equivalence to the classical problem of finding the Maximum Weight Matching (MWM) on a bipartite graph. This is a well-studied problem in classical graph theory and several well-known solutions exist. For the optimal assignment, we consider the Hungarian algorithm - which always achieves the maximal network utility — and compare its performance to a sub-optimal modified min-sum belief propagation algorithm. Through simulations, we demonstrate the negligible difference in performance between the two.