Research On Spectrum Sensing Technique In Wireless Sensor Network

The recent rapid growth of wireless communications has made the problem of spectrum utilization even more critical. This prompted cognitive radio (CR) to be an exciting emerging technology that has the potential of dealing with the stringent requirement and scarcity of the radio spectrum. Combining the adaptive spectrum sensing technology in CR with wireless sensor network (WSN) can reduce the cochannel interference and increase the channel occupancy efficiently. Therefore, spectrum sensing in WSN will be studied thoroughly in this thesis.Firstly, several typical techniques of spectrum sensing are introduced and their relative merits are compared in detail. Secondly, a spectrum sensing algorithm based on energy detection of complex signals is proposed according to the spectrum sensing demand under complex electromagnetic environment. The introduction of the central limit theorem makes the decision statistic approximate the normal distribution, thus it greatly simplifies the following calculative process and the theoretical deduction. The simulation results show that the proposed algorithm can obtain the probability of detection over 95% under -5 dB SNR condition, with the constant probability of false alarm 0.01. Thirdly, the hard and soft decision combining based on data fusion are also discussed. Computer simulations show that the latter exhibits much better performance than the former. Finally, the problem of detection under noise uncertainty is considered. Meanwhile, M2M4 algorithm used to estimate noise variance is analyzed and its effect on performance is simulated. In summary the algorithm proposed in this thesis with good performance is easy to realize and satisfies system requirements.