Experimental Research On Spectrum Sensing Based On Multi-node Co-location

Along with the rapid development of human society,communication is more and more widely used by people,at the same time,people put forward more and more high requirements on communication technology.But,spectrum resource is limited,ant it which as a prerequisite to improve its utilization is more and more become the research focus.This paper,a broad band will be divided into several narrow band spectrum by distributed thought,and the data is collected and transmitted by the wireless sensor network node.We also studied the multi-node co-location,and improved the classic APIT localization algorithm.In terms of spectrum sensing,multiple sets of low rate of sampling A/D are used and we recovery spectrum through the frequency estimation algorithm based on signal time delay and multiple rate,then,we improved these two kinds of frequency estimation.In the end,we also analyzed the noise threshold and research.The research work of this paper is as follows:1)Fourier transform theory and spectrum sensing technology is studied in detail,the discrete Fourier transform,fast Fourier transform,and main several kinds of spectrum sensing technology accordingly introduced,and we compared the advantages and disadvantages of several kinds of spectrum sensing technology;Introduced the overview,characteristics and application of wireless sensor network,and studied several classical node self-locating algorithms,and simulated and analyzed some of them;2)In the term of building hardware system architecture,a dual processor architecture is used in the this paper.Signal data acquisition and transmission is realized by using FPGA,MCU msp430 realizes multi-node co-location and time synchronization,and time domain signal collected is emulated and analyzed through the SignalTap II in the Quartus II.In terms of multi-node cooperative localization,we improves the classical APIT localization algorithm,and use the wireless transmission of radio frequency signal strength(RSSI)distance to realize node localization.In the end,we made simulation experiments of the relevant parameters;3)Due to the low rate signal sampled can cause spectrum aliasing,this paper studies the frequency estimation algorithm based on signal time delay and multiple rate,we realize frequency separation and calculation by constructing multiple sets of different sampling rate signal,and optimized and improved for existing problems.Then,we design a data fusion frequency estimation algorithm based on trust,also analysis the noise threshold and research,finally done simulation analysis and algorithm performance evaluation in the Matlab,obtained the good experimental effect.