Self-Organize Network Architecture for Multi-agent Cognitive Radio Systems

This paper presents self-organize network architecture for cognitive radio (CR). In order to increase the flexibility of CR, a division method and corresponding network keeping strategy is proposed in this paper, whose key part is finding the cutting edge in the interface graph. The TFCE (Tarjan-find-cutting-edge) algorithm is also given to find the cutting edge. On the basis of the division method a spectrum allocation model is introduced to research the effectiveness of the method for CR. The result shows the self-organize method could effectively add group´s flexibility without affecting spectrum allocation utility or increasing system traffic.