Efficient spectrum matching based on spectrum characteristics in cognitive radio systems

QoS is a very important issue in cognitive radio systems. We present and discuss spectrum matching algorithms to support QoS for cognitive radio users. This work is carried out in the context of a dynamic changing spectrum. In this paper, we propose efficient spectrum matching algorithms, which have low-complexity and quite good performance, based on statistical characteristics of spectrum bands. Those are classified random matching algorithm and greedy algorithm. The classified random matching algorithms would match the most suitable spectrum class to a proper user and randomly choose a spectrum within the class. And, using greedy algorithm, the best spectrum out of available spectrum bands will be chosen whenever a new user joins. The proposed algorithms provide QoS for cognitive radio systems through the reduction of spectrum handoff probability and improvement of fairness. Performance analysis and simulation results validate the efficiency of the proposed algorithms.