Title:MAPEL: Achieving Global Optimality for a Non-convex Wireless Power Control Problem

Abstract: Due to the complicated coupling and non-convexity introduced by interference, optimal joint SINR allocation and power control is difficult to achieve in wireless networks. Previous work in this area has mainly focused on the cases where the problem can be transformed into a convex optimization problem through proper change of variables. Unfortunately, the important problem of weighted throughput maximization (WTM) cannot be convexified using the existing approaches, and thus achieving its optimal performance has been an open problem. This paper proposed a centralized algorithm, MAPEL, which is guaranteed to globally converge to a global optimal solution of the weighted throughput optimization problem. The success of MAPEL algorithm is based on three key observations of the WTM problem: (1) the monotonicity of the objective function in terms of signal to interference-plus-noise ratio (SINR), (2) the objective function can be transformed into a product of exponentiated linear fraction functions, and (3) the feasible SINR region is always normal although not necessarily convex. Furthermore, MAPEL can trade off performance for convergence time by tuning its own approximation factor. Although a centralized algorithm, MAPEL provides an important benchmark for performance evaluation of existing or newly-proposed heuristic algorithms in this field. Several case studies on performance evaluations of state-of-art centralized and distributed algorithms are presented.