Research And Implementation Of Spectrum Sensing Technology Based On Multi-Core DSP

Spectrum sensing can make a corresponding judgment based on the parameter characteristic of the received signal.It is regarded as an essential and intricate operation in cognitive radio,and becomes a rapidly evolving area with various civilian and military applications.The free frequency band can be sensed without affecting the communication of authorized users by Spectrum sensing,which can lead to the recycling utilization of Spectrum,one can perform the improvement of the spectrum efficiency.A new spectrum sensing algorithm based on the SNR uncertainty is proposed in this thesis,which can achieve much better spectrum efficiency than the existing technique.In order to satisfy the hardware requirement of the proposed method,a multi-core DSP parallel signal processing board card scheme based on FPGA online management is designed and tested.This thesis can be generally divided into several categories:1?The accuracy and the computational complexity of the traditional method is investigated in detail.The single-cognitive node spectrum sensing can be easily affected by the shadow,which can lead to the terrible performance of spectrum sensing.The principles and characteristics of the distributed and centralized spectrum sensing technologies are introduced and analysed,and the fusion algorithms of centralized spectrum sensing are introduced and analyzed.2?The energy detection algorithm has lower computational complexity whereas leads to unsatisfactory performances.In contrast to the energy detection algorithm,the cyclostationary algorithm may not be efficient to implement but can lead to the optimal performance.Thus,a new spectrum sensing algorithm based on the SNR uncertainty is proposed in order to achieve better spectrum sensing performance.3?A multi-core DSP parallel signal processing board card scheme based on FPGA online management is proposed in this thesis.Firstly,the overall architecture diagram of the system is introduced and the hardware and software structure of the system is analysed in detail.Then,the hardware parts are tested systematically.Finally,task allocation,intercore communication and synchronous resource are implemented on a multi-core DSP to complete signal processing tasks.The design of multi-core DSP parallel signal processing board based on FPGA online management is tested and verified.The results demonstrate that the system can work normally.The modules of the proposed spectrum sensing algorithm are analyzed is detail and the algorithm is realized on the hardware platform.