Research On Cyclostationary Technology In Spectrum Sensing Of Cognitive Radio

Without interfering with the norm al communication of authorized users, Cognitive radio(CR) improves the spectrum resource utilization ratio, and reduces the cost of spectrum resources.This thesis mainly focused on cyclostationary based technology in spectrum sensing of CR. Diff erent characteristic of spectrum sensing methods were summarized in the paper a nd cyclostationary base d spectrum sensing technology was chosen as concentration. In the paper, different proven m ethods of plotting Cyclic Spectrum Correlation Function (CSCF) were used. The basic theory of signal detection of Cyclostationary Autocorrelation Function(CAF) based method was analyzed. Performance simulation and com parison with Ene rgy Detection were also made.In order to overcom e the shadow effect that happened in single user detection, and to im prove the perfor mance of detecti on, this paper com bined cyclostationary based detection with collaborative sensing. Performance comparison were m ade between different traditional schemes of soft combining and hard combining. The scheme we use takes advantage of SN R inform ation and Chair-Varshney Rule. Simulation reveals that our scheme can im prove spectrum sensing perform ance without increasing data transmission in data fusion centre. Experiments also show that the degree of the im provement is related to the SNR distribution of the cooperative secondary users.Finally, in order to facilitate the SCF estimation algorithm on special integrated circuit, We simulated FAM algorithm in fixed-point C language and achieved the desired results. At the sam e time as th e core algorithm of FAM, fixed-point FFT algorithm was carefully studi ed, and sim ulations were m ade. Though program in general computer can’t be put into use in hardware platform directly, but it can also give thereotical ditection to hardware implementation.