A Novel Low-Complexity Dynamic Frequency Selection Algorithm for Cognitive Radios

Abstract

In this paper, a non-cooperative distributed frequency selection algorithm is proposed for cognitive radio networks in which in addition to interference minimization, the minimization of implementation complexity and probability of interruption of secondary users are also addressed. The convergence and interference reduction property of the algorithm is also demonstrated using game theoretic concepts. In order to evaluate the performance of the algorithm, the associated optimization problem is formulated into a variant of graph colouring problem, and a centralized algorithm is also proposed as an approximation of the optimal solution. Finally, simulation results demonstrate that the proposed distributed algorithm achieves acceptable performance in comparison with earlier schemes.