Power and modulation assignment via Perron-root optimization for interference limited systems

The maximization of the total sum rate depends on the proper power and modulation assignment. The feasibility of such resource assignment is susceptible to the set of links that are attempted to be scheduled. In this paper we address the problem of maximizing the sum rate while guaranteeing resource assignment feasibility in interference limited wireless systems. Unlike the current literature, the allowed signal-to-interference-plus-noise ratio is constrained to take values from a finite set associated with a finite number of available modulations. Therefore, we present suboptimal but efficient algorithms to solve the joint user selection and resource (power and modulation) assignment, which is a combinatorial problem. The feasibility of the resource assignment is verified by a criterion based on the Perron-Frobenius theory whilst the optimization of the resource allocation is achieved either by user-removal like techniques or by a novel criterion derived from the Perron-Frobenius theory. We provide numerical results to confirm the efficiency of our resource allocation algorithms compared to the optimal resource allocation.