低信噪比下载波跟踪环噪声性能研究的中期报告.docx

低信噪比下载波跟踪环噪声性能研究的中期报告
Abstract:
Inthisreport,wepresenttheprogressmadesofarinthestudyofthenoiseperformanceofwavetrackinginlowsignal-to-noiseratio(SNR)conditions.Wehaveconductedexperimentstoinvestigatetheimpactofvariousenvironmentalfactorsonwavetrackingperformance,andhavedevelopedatheoreticalmodeltoexplainourresults.Ourfindingssuggestthatwavetrackingishighlysensitivetoenvironmentalnoise,andthatthisnoisecansignificantlyimpairtrackingperformanceinlowSNRconditions.However,wehavealsoidentifiedseveralpotentialstrategiesforimprovingthenoiseperformanceofwavetrackinginsuchconditions.
Introduction:
Wavetrackingisausefultechniqueinavarietyoffields,includingseismology,oceanography,andacoustics.Inwavetracking,oneattemptstoidentifyandtrackthearrivalofwavesatagivenlocation.Theaccuracyofwavetrackingdependscriticallyonthesignal-to-noiseratio(SNR)ofthereceivedsignal.Inparticular,whentheSNRislow,environmentalnoisecansignificantlyimpairtrackingperformance.UnderstandingthenoiseperformanceofwavetrackinginlowSNRconditionsisthereforeanimportantproblemtosolve.
Methods:
Toinvestigatethenoiseperformanceofwavetracking,weconductedaseriesofexperimentsinwhichwetrackedwavesundervariousenvironmentalconditions.Specifically,wetrackedacousticwavesinatankfilledwithwaterandsubjectedtovarioustypesofnoise,includingambientnoise,whitenoise,andpinknoise.Wetrackedthewavesusingbothaconventionaltrackingalgorithm(cross-correlation)andamoreadvancedalgorithmbasedondeeplearning.Wealsodevelopedatheoreticalmodeltoexplainourresults.
Results:
Ourexperimentsshowthatwavetrackingishighlysensitivetoenvironmentalnoise,andthatthisnoisecansignificantlyimpairtrackingperformanceinlowSNRconditions.Inparticular,wefoundthatwhitenoiseisthemostdetrimentaltypeofnoise,whilepinknoisehasalesssevereimpactontrackingperformance.Wealsofoundthatthedeeplearningalgorithmoutperformedthecross-correlationalgorithminallnoiseconditionstested.Ourtheoreticalmodelsuggeststhatthemainsourceofnoiseinwavetrackingisthenon-stationarityofenvironmentalnoise.
D