Research Of Periodic Non-uniform Sampling Based Wideband Spectrum Sensing

With the rapid development of wireless communication technologies and the sharp increasement in the number of mobile users,the problem of wireless spectrum scarcity has become increasingly serious.Cognitive radio(CR),an intelligent radio technology can effectively improve the utilization efficiency of spectrum resources.Because it can continuely perceive external environment to detect and use the untapped primary users'free spectrum.In the system of wideband spectrum sensing of cognitive radio,the primary users have lower utilization of spectrum resources.So they only take up a small amount of spectrum resources.That leads to primary users' sparse characteristics in the frequency domain.If we use uniform sampling,the Analog-to-Digital Converter(ADC)must have high sampling rate,which will generate a lot of data redundancy.However,Periodic Non-uniform Sampling(PNS)technology can make use of the sparse characteristics of primary users in frequency domain.It can directly scan the wideband signal in low sampling rate that is well below Nyquist sampling rate.PNS technology can not only reduce the cognitive users' burden on sampling,but can achieve rapid and efficient wideband spectrum sensing.The main research contributions are as follows:(1)We make use of Depth-First Search(DFS)algorithm to search the optimal sampling sequence.(2)We use the optimal sampling sequence for scanning wideband signal.Then applie the unbiased Power Spectral Density(PSD)algorithm to calculate the PSD of PNS points so that we can determine the active state of the primary user signal.(3)We can detect small signals' power spectrum which is submerged by the big primary users' power spectrum with the use of PNS technology and Least Squares(LS)method so that we can avoid missed diagnosis of primary user signals..(4)We design and implement PNS system in the core of FPGA and STM32.The thesis includes the design and achievement of PNS clock function,external trigger function,transmission function of storage of sampling and data.We introduce the building,serial transmission,testing and key technologies of sampling system.In the thesis,if we know the range of wideband,we can make use of PNS technology to scan the wideband signal at a lower average sampling rate in the absence of prior knowledge(such as spectral position)of the primary user signals and calculate its PSD.Then we can identify submerged small primary user signals' power spectrum by the big primary users' power spectrum with the use of LS method.The PNS system designed and implemented in this paper has important value in practical applications.