Abstract: When some primary users re-emergence，the secondary users must immediately give up the spectrum belonging to the primary users.At this point there is a problem on how to maintain the existence of the Secondary Users’Link（SUL）.The redundant sub channels model based on CORVUS system use another X redundant sub channels besides the original N sub channels for the communications among secondary users.As long as the interfered sub channels during the transmission are less than X，the receiver can restore the correct information from the N+X sub channels，thus solving the SUL maintaining problem.However so far，there is no effective way to solve the optimal N and X.This paper models the SUL maintaining problem as a multiobjective optimization problem，develops a new SUL maintaining algorithm called SULEA based on a multiobjective evolutionary algorithm.SULEA can dynamically select the fitness functions according to the user’s different service need，so as to solve the optimal N and X，and the results of Matlab simulation experiment testify the validity of SULEA.

Key words: cognitive radio, multiobjective evolutionary algorithm, Secondary Users’Link（SUL） maintaining problem, redundant sub channels model, SUL Maintaining Algorithm Based on Multiobjective Evolutionary Algorithm（SULEA）

摘要： 当某个一级用户重新出现时，二级用户必须马上让出属于此一级用户的频谱，此时就存在着如何维持二级用户链接（Secondary Users’Link，SUL）的问题。基于CORVUS系统的“冗余子信道”模型在SUL原有的N个子信道基础上再使用X个冗余子信道进行二级用户之间的通信。只要传输过程中受干扰的子信道个数少于X，接收端就可以从N+X个子信道中恢复出正确信息来，从而解决了SUL维持的问题。但尚未有一种有效的办法以求解最优的N和X。将SUL维持问题建模为一个多目标优化问题，提出了一种基于多目标遗传算法的SUL维持算法SULEA。SULEA能够根据不同的用户服务需求动态地选择适应度函数来求解最优的N和X，Matlab实验证明了SULEA的正确性和有效性。

关键词: 认知无线电, 多目标遗传算法, 二级用户链接（SUL）维持问题, 冗余子信道模型, 基于多目标遗传算法的SUL维持算法（SULEA）