Research On Bandpass Signal Detection Method Based On Power Spectrum

In the background of strong noise and interference,the detection of bandpass signal is always a difficult problem in the field of engineering application.With the popularization of wireless communication,various electromagnetic devices are more and more widely used,which makes the detection of bandpass signal more and more difficult.In the complex electromagnetic environment,the traditional signal detection method has been difficult to meet the requirements of real-time application system.Therefore,this paper makes an in-depth study on the bandpass signal detection method.The main researches of this paper can be concluded as follows:1.The classical energy detection algorithm and template matching algorithm are studied.Because noise has great influence on the energy detection algorithm,and noise cannot be avoided in the actual system,this paper introduces the energy detection algorithm based on noise uncertainty in detail.When the noise is uncertain,single threshold detection may cause false alarm,so double threshold detection is proposed to reduce false alarm probability.Under the condition of knowing the prior knowledge of the signal,in order to make full use of the prior knowledge of the signal to improve the detection performance,the template detection algorithm is analyzed in detail,and its principle and specific implementation flow are introduced.In this paper,the performance of these two algorithms in complex environment is analyzed through the power spectrum data collected in practice,and the advantages and disadvantages of these two classical detection algorithms are summarized.2.Due to the poor detection performance of the two classical detection algorithms mentioned above in the complex electromagnetic environment,this paper introduces a new type of bandpss signal detection algorithm--morphological filtering detection algorithm from imagology.It makes full use of the bandwidth characteristics of the known power spectrum,and filters out all the interference signals which are smaller than the target signal bandwidth to achieve the purpose of signal detection.In this paper,the principle of morphological detection algorithm is first introduced,and the selection basis of structural elements of key points of the algorithm is explained in detail through simulation.Then the implementation flow of the algorithm is given,and the feasibility of the algorithm is verified by simulation and measured power spectrum data,and its performance in complex environment is analyzed.Finally,the shortcomings of the algorithm are improved,and the improved algorithm can filter out the interference signals with smaller bandwidth than the target signal and the interference signals with larger bandwidth than the target signal,thus reducing the false alarm rate.3.Because the classic signal detection algorithms use a fixed threshold to estimate the background value,If the threshold is too large,it will cause a false alarm and If it is too small,it will cause a false alarm.To solve the above problem,this paper proposes a frequency domain constant false alarm detection algorithm whose threshold value can dynamically change with the background value.In this paper,the mean constant false alarm detector and the ordered constant false alarm detector are analyzed in detail,and the principle is introduced first.Then the feasibility of the frequency domain is verified by the specific process and actual power spectrum data,and its performance in complex environment is analyzed.Finally,according to the measured results,improvements are made to reduce the probability of false alarms.