中文摘要：

机会式频谱接入技术虽能够有效提升认知无线网络频谱利用率,但在保障认知用户服务质量（quality of service,QoS）以及频谱资源有效利用率方面存在不足.为了提高认知用户QoS,提升认知无线网络系统性能,提出一种基于区间Mamdani模糊推理的认知无线网络频谱迁移方案.通过引入预判决方法减少不必要的数据量和系统开销.综合考虑待迁移频谱的频谱占用率以及链路维持率,并利用模糊推理计算出频谱迁移度,指导认知用户迁移至最优的频谱空穴.为了缩短模糊推理时长,提出区间Mamdani模糊推理方法.仿真结果表明,该方案能够降低认知用户业务传输的强制中断率、重传率以及频谱迁移次数,在维持较高系统吞吐量的同时,提高认知无线网络频谱资源的有效利用率.

英文摘要：

The existing opportunistic spectrum access technologies can improve spectrum resource utilization significantly, but they still have some shortcomings in guaranteeing QoS of secondary users and effective utilization probability of spectrum resource in cognitive radio networks. In order to improve QoS of secondary users in cognitive radio networks, and promote the real performance of cognitive radio networks, a new concept named spectrum migration is introduced, and a novel spectrum migration scheme based on interval Mamdani fuzzy inference is put forward. In the spectrum migration scheme, unnecessary treated data and system load can be reduced by applying pre-decision method. In addition, spectrum occupation rate and link maintenance rate of licensed spectrum are comprehensively considered as spectrum migration factors. Then, spectrum migration degree can be calculated by fuzzy inference to guide secondary users to the optimal spectrum holes. To shorten inference time, interval Mamdani fuzzy inference is proposed based on Mamdani fuzzy inference, and the twice judgments are utilized to reduce the complexity of the spectrum migration process. Simulation results show that the scheme can decrease forced termination probability, service retransmission probability and spectrum migration times of secondary users＇ service transmission, maintain a higher system throughput and improve the effective utilization probability of cognitive radio networks spectrum resource at the same time.