Efficient joint power and admission control in underlay cognitive networks using Benders' decomposition method

Abstract In this paper, the joint power and admission control (JPAC) problem in cognitive radio networks is studied. This problem is decomposed into two subproblems, by utilizing the Benders’ decomposition theory, which are efficiently addressed. Specifically, the former is a simple linear optimization problem which a closed-form expression for its optimal solution is derived, and the latter is solved via an iterative distributed power control algorithm. Furthermore, we use the outcomes of the decomposed two subproblems to propose a sequential searching JPAC algorithm with a novel removal metric whereas minimal number of SUs are sequentially removed. In an infeasible system, where the minimum target signal-to-interference-plus-noise-ratios (SINRs) of all primary and secondary users are not simultaneously reachable, our proposed JPAC algorithm guarantees protecting all primary users while the maximal number of SUs are admitted and supported with their target SINRs. Not only our proposed algorithm does converge to an equilibrium, but also outperforms existing algorithms in terms of average outage ratio and average aggregate power, as demonstrated through the extensive simulations.