Research On Unaided GNSS Signal Detection

The new applications of Global Navigation Satellite System (GNSS) have higherrequirement on signal acquisition when there are no aided facilities or aided facilitiescannot be obtained. So unaided GNSS signal detection is important to be researchedfurther. Theoretically, the mathematical model of the unaided GNSS signal detectionshould be set up, and its statistical distribution should be explored. To improve the postprocessing signal to noise ratio, the unaided weak signal detection requires long coherentintegration time and many noncoherent integration steps. The disadvantages of doing theseare the increase of the computation burden and the acquisition time. Thus it’s necessary toimprove the current acquisition method and develop new acquisition methods to reduce thecomputation burden and the acquisition time. On the aspect of receiver hardware, thenumber of correlators needed by the average acquisition method is equal the size of thematrix combined by the Doppler and code searching space. A new concept calledcompressive sensing is adopted to do GPS normal and weak signal acquisition. Thecompressive sensing acquisition method balances the number of correlators and theacquisition sensitivity, so it makes reducing hardware correlators possible. This thesismainly contains the following aspects:(1) Both the intermediate frequency (IF) and baseband GPS signal detection modelswhich applies the hypothesis testing analysis are set up. The statistic expressions, thedetection performance (probability of false alarm and probability of detection) and theunknown parameters are fully analyzed under4scenarios: the Doppler frequency and codeare known, the Doppler frequency and code are unknown, combining the noncoherentintegration, and weak signal.(2) For weak signals, the performance and expressions in terms of detection and falsealarm probabilities for4different noncoherent integration alternatives (named Alternative1, Alternative2, Alternative3and Alternative4) are investigated and derived, respectively.The theoretical analysis and simulation analysis show that Alternative2is nearly3dBmore sensitive than Alternative1. And the alternatives referring to acquisition sensitivityfrom low to high are Alternative2, Alternative3, Alternative4and Alternative1.(3) Compared with the Fast Modified Double-Block Zero-Padding (FMDBZP)algorithm applying the Noncoherent Integration Alternative2, the proposed ImprovedFMDBZP (IFMDBZP) algorithm can reduce the computation calculations and savememory space without suffering any loss. And the longer the coherent integration time is,the more benefit one can gain from the proposed IFMDBZP algorithm. So the performanceof IFMDBZP is the same with FMDBZP using Noncoherent Integration Alternative2, andbetter than FMDBZP using Noncoherent Integration Alternative1, Alternative3orAlternative4. (4) Compressive sensing theory is introduced to do GPS signal acquisition, whichbreaks through the processing data length limitation and the data do not have to be limitedin one or a few navigation data bit length. For the first time the impact of the compressivesensing on acquisition sensitivity is analyzed, and the affection of noise folding caused bycompressive sensing is explored. And the correct simulation results and conclusions areobtained.