摘要
为了提高频谱感知的整体性能,基于链路层多认知用户集中式协作感知思想,提出一种带缓冲区的双周期n步串行协作感知机制。该机制利用多认知用户分时、分段协作提高频谱感知效率,建立频谱池缩短被中断用户切换延迟时间,使用离散马尔可夫模型对感知参数建模,通过求解最优搜索步长和双感知周期比,提高认知用户频谱感知性能和QoS。仿真实验结果显示,该算法在感知效率、被迫中断概率和中断时间方面均优于随机搜索和传统串行搜索策略。
Spectrum sensing is identified as the fundamental issue in cognitive radio networks（CRNs）.A doubly periodic n-serial combined sensing mechanism with spectrum pool based on combined spectrum sensing is proposed to improve the overall sensing performance.The spectrum sensing efficiency is improved using different cognitive users time division and section division coope-rative sensing,spectrum pool is established to shorten the delay due in channel switching,and the discrete Markov parameter optimization model is set up.The cognitive users spectrum sensing performance and QoS are improved by solving the optimal channel section number and double sensing period ratio.The simulation results show that this algorithm is obviously superior to the random search and traditional serial search strategy at sensing efficiency,interrupted rate and interrupted time.
出处
《计算机工程与设计》
CSCD
北大核心
2013年第8期2662-2667,2694,共7页
Computer Engineering and Design
基金
国家自然科学基金项目(61074058)
国家教育部博士点基金项目(20090162120068)
广东省自然科学基金项目(S2011040004769)
关键词
认知无线电
频谱感知
频谱池
协作感知
马尔可夫决策过程
cognitive radio
spectrum sensing
spectrum pool
cooperation spectrum sensing
Markov decision process
