基于频谱预测和频谱分割的吞吐量优化  

Optimization of throughput based on spectrum prediction and spectrum segmentation

作　　者：李红[1] 齐丽娜[1] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003

出　　处：《南京邮电大学学报（自然科学版）》2016年第2期60-64,共5页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家重点基础研究发展计划(973计划)(2013CB329005);国家自然科学基金(61471201)资助项目

摘　　要：在认知无线电网络中,进行周期性的频谱感知时,次用户需要在感知阶段等待,这就减少了次用户的吞吐量。另外当次用户在多条信道中随机选择信道感知时,感知结果为占用的概率较高,造成次用户的吞吐量较低。为了提高次用户的通信量,文中结合频谱分割和频谱预测技术,重新设计次用户的帧结构,先将主用户的频谱划分为两个子频带,然后在一个子频带中加入频谱预测功能,使次用户只在预测为空闲的信道中选择感知信道,从而提高感知为空闲的概率,进而提高系统的吞吐量。仿真结果表明,该方法可以有效的提高次用户的吞吐量。In a cognitive radio（ CR） network,when primary user（ PU） spectrum is periodically sensed,the secondary user（ SU） needs to wait during sensing interval,thus causing the interruption in SU transmission and significant reduction in the achievable throughput. In addition,when SU randomly selects channels for the spectrum,the probability of the busy channel is high,resulting in the low throughput.Based on the spectrum prediction and the spectrum segmentation,a new frame structure is designed. Firstly,the band of the channel is divided into two subbands. Then,the spectrum prediction slot is added into the subband for the spectrum sensing. SU only senses the channel status of the slot which is predicted to be idle,so the throughput is improved. Simulation results show that the proposed method can improve the SU throughput.

关 键 词：频谱感知 认知无线电网络 吞吐量 频谱分割 频谱预测 

分 类 号：TN92[电子电信—通信与信息系统]