基于GNU Radio和USRP的能量检测频谱感知技术研究

发布时间：2019-03-08 13:44

【摘要】：随着5G技术的到来,物联网技术的出现,以及众多基于无线电的创新性服务,已经拥挤的无线电资源变得更加紧张。因此,不仅需要提供更多的频谱,而且使现有频谱的使用变得更有效率也是至关重要的。认知无线电技术可以有效的解决这一问题,本文主要借助软件无线电平台USRP,研究基于能量检测的频谱感知技术。GNU Radio是一套开源的软件系统,内含丰富的信号处理模块,可通过USRP连接PC进行一系列的无线电开发操作,方便快捷的设计通信系统模型。本文主要工作和创新点如下:第一,在USRP平台上实现了基于Python框架的能量检测频谱感知技术。充分利用了GNU Radio的功能,消除了第三方处理软件的需求,大大减少了计算开销,介绍了应用经典的FFT模块进行快速傅里叶变换实现频谱感知功能的流程。第二,提出并实现了基于USRP的两阶段频谱感知技术方案。分别是粗扫描阶段和细扫描阶段。粗扫描阶段可以实时检测50MHz-2200MHz频段的频谱占用情况,发现频谱空洞及目标信号。细扫描阶段可以在指定的频段进行精确地扫描,检测信号的中心频率和带宽以及观测信号的频谱图。利用粗扫描过程进行快速检测发现的空闲频段进一步通过细扫描过程可提高检测精度。在细扫描阶段结束后,作为认知用户的USRP可以自行选择适当的频段发送数据包进行频谱接入。通过比较数据包的正确接收率,经过频谱感知过程选择合理的频段进行接入会大大降低通信过程的误码率。第三,设计了有主用户干扰下的USRP图像传输实验,比较了基于不同的频谱感知方式选择频段传输图像的效果,计算了图像文件的正确接收率。基于粗扫描和细扫描两阶段的频谱感知方案不仅感知精度高,而且节省了感知的时间,因此本方案具有最优的检测效果。
[Abstract]:With the advent of 5G technology, the emergence of Internet of things technology, and many innovative radio-based services, the already crowded radio resources have become more tense. Therefore, it is essential not only to provide more spectrum, but also to make the existing spectrum more efficient. Cognitive radio technology can solve this problem effectively. In this paper, we mainly use software radio platform USRP, to study spectrum sensing technology based on energy detection. GNU Radio is an open source software system with rich signal processing modules. A series of radio development operations can be carried out by connecting PC with USRP, and the communication system model can be designed conveniently and quickly. The main work and innovations of this paper are as follows: firstly, the energy sensing spectrum sensing technology based on Python framework is implemented on the USRP platform. Taking full advantage of the function of GNU Radio, the requirement of the third party processing software is eliminated, and the computing overhead is greatly reduced. The flow of realizing spectrum sensing function by using the classical FFT module is introduced in this paper. Secondly, a two-stage spectrum sensing scheme based on USRP is proposed and implemented. They are coarse scan phase and fine scan phase respectively. In the coarse scanning phase, the spectrum occupation of 50MHz-2200MHz band can be detected in real time, and the spectrum cavity and target signal can be found. The fine-scan phase can be accurately scanned in the specified frequency band to detect the central frequency and bandwidth of the signal as well as the spectrum map of the observed signal. The detection accuracy can be improved by fine scanning in the idle frequency band, which is found by coarse scanning process. At the end of the fine-scan phase, the USRP, as a cognitive user, can select the appropriate frequency band to send packets for spectrum access. By comparing the correct reception rate of the data packet, the error rate of the communication process can be greatly reduced by selecting a reasonable frequency band through the spectrum sensing process. Thirdly, the experiment of USRP image transmission with host user interference is designed. The effect of selecting frequency band based on different spectrum sensing methods is compared, and the correct recovery rate of image file is calculated. The spectrum sensing scheme based on coarse scan and fine scan not only has high sensing precision, but also saves sensing time, so this scheme has the best detection effect.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925

【参考文献】