Joint optimization for downlink resource allocation in cognitive radio cellular networks

Abstract

This paper takes into account the uncertainty of the primary users' locations and transmission power in designing an optimal downlink scheduling scheme for cognitive radio cellular networks (CogCells). Localization technique is exploited to estimate the position and transmission power of the primary user (PU) transmitting on specific channel. The objective of our scheduling scheme is to maximize the downlink average throughput for CogCells without causing harmful interference to PUs. This paper models the problem as a mixed integer nonlinear programming (MINLP) problem, which has exponential complexity by traditional direct search method. An efficient joint channel assignment and power control scheme based on dual decomposition method is proposed. Firstly, the dual optimization problem is decomposed into K independent subproblems of channel assignment. Karush-Kuhn-Tucker (KKT) conditions are then applied to find the optimal user for a specific channel. Secondly, ellipsoid method is applied to update the dual variables and find the optimal solution for the primal problem. Numerical results demonstrate the effectiveness of the proposed scheme.